Training in progress, epoch 1, checkpoint
Browse files
checkpoint-1469/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 383295784
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41e784c8c66a8a24b891f9e7cb9e5c291d1946482d489927bdba0617899678a5
|
| 3 |
size 383295784
|
checkpoint-1469/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:930a7b2f85ed02cbfb8792d0ca762b70ae705518d7f4c19b05ed96f62b8cdbd0
|
| 3 |
+
size 184533306
|
checkpoint-1469/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14308
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b8c11102d4b341eb099e507d80d3c782f420154e1604e2d7f85defeb24e33f2
|
| 3 |
size 14308
|
checkpoint-1469/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2ec84729d9ec9e310defe44d38dfd41ab0bf0bd1f9f56fc1eae0c101f6d849b
|
| 3 |
size 1064
|
checkpoint-1469/trainer_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint":
|
| 4 |
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 1469,
|
|
@@ -8,2068 +8,537 @@
|
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.0034036759700476512,
|
| 13 |
-
"grad_norm": 24747.943359375,
|
| 14 |
-
"learning_rate": 1.7018379850238258e-08,
|
| 15 |
-
"loss": 2.0794,
|
| 16 |
-
"step": 5
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.0068073519400953025,
|
| 20 |
-
"grad_norm": 40961.8125,
|
| 21 |
-
"learning_rate": 3.4036759700476516e-08,
|
| 22 |
-
"loss": 2.0796,
|
| 23 |
-
"step": 10
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.010211027910142955,
|
| 27 |
-
"grad_norm": 36985.2734375,
|
| 28 |
-
"learning_rate": 5.105513955071478e-08,
|
| 29 |
-
"loss": 2.0793,
|
| 30 |
-
"step": 15
|
| 31 |
-
},
|
| 32 |
{
|
| 33 |
"epoch": 0.013614703880190605,
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss": 2.0797,
|
| 37 |
-
"step": 20
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.01701837985023826,
|
| 41 |
-
"grad_norm": 32935.01953125,
|
| 42 |
-
"learning_rate": 8.50918992511913e-08,
|
| 43 |
"loss": 2.0802,
|
| 44 |
-
"step":
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.02042205582028591,
|
| 48 |
-
"grad_norm": 17762.544921875,
|
| 49 |
-
"learning_rate": 1.0211027910142956e-07,
|
| 50 |
-
"loss": 2.079,
|
| 51 |
-
"step": 30
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.023825731790333562,
|
| 55 |
-
"grad_norm": 28104.220703125,
|
| 56 |
-
"learning_rate": 1.1912865895166782e-07,
|
| 57 |
-
"loss": 2.0786,
|
| 58 |
-
"step": 35
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.02722940776038121,
|
| 62 |
-
"grad_norm":
|
| 63 |
-
"learning_rate":
|
| 64 |
-
"loss": 2.
|
| 65 |
"step": 40
|
| 66 |
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.03063308373042886,
|
| 69 |
-
"grad_norm": 26094.353515625,
|
| 70 |
-
"learning_rate": 1.5316541865214431e-07,
|
| 71 |
-
"loss": 2.0791,
|
| 72 |
-
"step": 45
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.03403675970047652,
|
| 76 |
-
"grad_norm": 45408.37109375,
|
| 77 |
-
"learning_rate": 1.701837985023826e-07,
|
| 78 |
-
"loss": 2.079,
|
| 79 |
-
"step": 50
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.037440435670524165,
|
| 83 |
-
"grad_norm": 29806.216796875,
|
| 84 |
-
"learning_rate": 1.8720217835262084e-07,
|
| 85 |
-
"loss": 2.0794,
|
| 86 |
-
"step": 55
|
| 87 |
-
},
|
| 88 |
{
|
| 89 |
"epoch": 0.04084411164057182,
|
| 90 |
-
"grad_norm":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 60
|
| 94 |
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.04424778761061947,
|
| 97 |
-
"grad_norm": 36855.83984375,
|
| 98 |
-
"learning_rate": 2.2123893805309737e-07,
|
| 99 |
-
"loss": 2.079,
|
| 100 |
-
"step": 65
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.047651463580667124,
|
| 104 |
-
"grad_norm": 24015.33984375,
|
| 105 |
-
"learning_rate": 2.3825731790333565e-07,
|
| 106 |
-
"loss": 2.0788,
|
| 107 |
-
"step": 70
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.05105513955071477,
|
| 111 |
-
"grad_norm": 26997.51953125,
|
| 112 |
-
"learning_rate": 2.552756977535739e-07,
|
| 113 |
-
"loss": 2.079,
|
| 114 |
-
"step": 75
|
| 115 |
-
},
|
| 116 |
{
|
| 117 |
"epoch": 0.05445881552076242,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate":
|
| 120 |
-
"loss": 2.0787,
|
| 121 |
-
"step": 80
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.057862491490810075,
|
| 125 |
-
"grad_norm": 31673.52734375,
|
| 126 |
-
"learning_rate": 2.893124574540504e-07,
|
| 127 |
"loss": 2.0791,
|
| 128 |
-
"step":
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.06126616746085772,
|
| 132 |
-
"grad_norm": 31619.5703125,
|
| 133 |
-
"learning_rate": 3.0633083730428863e-07,
|
| 134 |
-
"loss": 2.0785,
|
| 135 |
-
"step": 90
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.06466984343090537,
|
| 139 |
-
"grad_norm": 29396.28125,
|
| 140 |
-
"learning_rate": 3.233492171545269e-07,
|
| 141 |
-
"loss": 2.0783,
|
| 142 |
-
"step": 95
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.06807351940095303,
|
| 146 |
-
"grad_norm":
|
| 147 |
-
"learning_rate":
|
| 148 |
-
"loss": 2.0785,
|
| 149 |
-
"step": 100
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.07147719537100068,
|
| 153 |
-
"grad_norm": 28501.669921875,
|
| 154 |
-
"learning_rate": 3.5738597685500344e-07,
|
| 155 |
-
"loss": 2.079,
|
| 156 |
-
"step": 105
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.07488087134104833,
|
| 160 |
-
"grad_norm": 31831.615234375,
|
| 161 |
-
"learning_rate": 3.744043567052417e-07,
|
| 162 |
-
"loss": 2.079,
|
| 163 |
-
"step": 110
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.07828454731109598,
|
| 167 |
-
"grad_norm": 24956.236328125,
|
| 168 |
-
"learning_rate": 3.9142273655547994e-07,
|
| 169 |
"loss": 2.0793,
|
| 170 |
-
"step":
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.08168822328114364,
|
| 174 |
-
"grad_norm":
|
| 175 |
-
"learning_rate":
|
| 176 |
-
"loss": 2.0794,
|
| 177 |
-
"step": 120
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.08509189925119129,
|
| 181 |
-
"grad_norm": 23309.07421875,
|
| 182 |
-
"learning_rate": 4.254594962559565e-07,
|
| 183 |
-
"loss": 2.0785,
|
| 184 |
-
"step": 125
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.08849557522123894,
|
| 188 |
-
"grad_norm": 29537.3046875,
|
| 189 |
-
"learning_rate": 4.4247787610619474e-07,
|
| 190 |
-
"loss": 2.0786,
|
| 191 |
-
"step": 130
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.09189925119128659,
|
| 195 |
-
"grad_norm": 32017.046875,
|
| 196 |
-
"learning_rate": 4.5949625595643294e-07,
|
| 197 |
"loss": 2.0793,
|
| 198 |
-
"step":
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.09530292716133425,
|
| 202 |
-
"grad_norm":
|
| 203 |
-
"learning_rate":
|
| 204 |
-
"loss": 2.
|
| 205 |
"step": 140
|
| 206 |
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.0987066031313819,
|
| 209 |
-
"grad_norm": 21977.302734375,
|
| 210 |
-
"learning_rate": 4.935330156569096e-07,
|
| 211 |
-
"loss": 2.0784,
|
| 212 |
-
"step": 145
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.10211027910142954,
|
| 216 |
-
"grad_norm": 25111.359375,
|
| 217 |
-
"learning_rate": 5.105513955071477e-07,
|
| 218 |
-
"loss": 2.079,
|
| 219 |
-
"step": 150
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.10551395507147719,
|
| 223 |
-
"grad_norm": 29474.87890625,
|
| 224 |
-
"learning_rate": 5.275697753573861e-07,
|
| 225 |
-
"loss": 2.0784,
|
| 226 |
-
"step": 155
|
| 227 |
-
},
|
| 228 |
{
|
| 229 |
"epoch": 0.10891763104152484,
|
| 230 |
-
"grad_norm":
|
| 231 |
-
"learning_rate":
|
| 232 |
-
"loss": 2.
|
| 233 |
"step": 160
|
| 234 |
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.1123213070115725,
|
| 237 |
-
"grad_norm": 32028.310546875,
|
| 238 |
-
"learning_rate": 5.616065350578626e-07,
|
| 239 |
-
"loss": 2.0778,
|
| 240 |
-
"step": 165
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.11572498298162015,
|
| 244 |
-
"grad_norm": 41068.41796875,
|
| 245 |
-
"learning_rate": 5.786249149081008e-07,
|
| 246 |
-
"loss": 2.0786,
|
| 247 |
-
"step": 170
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.1191286589516678,
|
| 251 |
-
"grad_norm": 27272.986328125,
|
| 252 |
-
"learning_rate": 5.956432947583391e-07,
|
| 253 |
-
"loss": 2.0788,
|
| 254 |
-
"step": 175
|
| 255 |
-
},
|
| 256 |
{
|
| 257 |
"epoch": 0.12253233492171545,
|
| 258 |
-
"grad_norm":
|
| 259 |
-
"learning_rate":
|
| 260 |
-
"loss": 2.
|
| 261 |
"step": 180
|
| 262 |
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.1259360108917631,
|
| 265 |
-
"grad_norm": 24603.845703125,
|
| 266 |
-
"learning_rate": 6.296800544588156e-07,
|
| 267 |
-
"loss": 2.0782,
|
| 268 |
-
"step": 185
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.12933968686181074,
|
| 272 |
-
"grad_norm": 29665.525390625,
|
| 273 |
-
"learning_rate": 6.466984343090538e-07,
|
| 274 |
-
"loss": 2.0782,
|
| 275 |
-
"step": 190
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.13274336283185842,
|
| 279 |
-
"grad_norm": 36893.68359375,
|
| 280 |
-
"learning_rate": 6.637168141592922e-07,
|
| 281 |
-
"loss": 2.0784,
|
| 282 |
-
"step": 195
|
| 283 |
-
},
|
| 284 |
{
|
| 285 |
"epoch": 0.13614703880190607,
|
| 286 |
-
"grad_norm":
|
| 287 |
-
"learning_rate":
|
| 288 |
-
"loss": 2.
|
| 289 |
"step": 200
|
| 290 |
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.13955071477195372,
|
| 293 |
-
"grad_norm": 25682.04296875,
|
| 294 |
-
"learning_rate": 6.977535738597687e-07,
|
| 295 |
-
"loss": 2.0776,
|
| 296 |
-
"step": 205
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.14295439074200136,
|
| 300 |
-
"grad_norm": 25126.154296875,
|
| 301 |
-
"learning_rate": 7.147719537100069e-07,
|
| 302 |
-
"loss": 2.0781,
|
| 303 |
-
"step": 210
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.146358066712049,
|
| 307 |
-
"grad_norm": 30238.875,
|
| 308 |
-
"learning_rate": 7.317903335602451e-07,
|
| 309 |
-
"loss": 2.0779,
|
| 310 |
-
"step": 215
|
| 311 |
-
},
|
| 312 |
{
|
| 313 |
"epoch": 0.14976174268209666,
|
| 314 |
-
"grad_norm":
|
| 315 |
-
"learning_rate":
|
| 316 |
-
"loss": 2.
|
| 317 |
"step": 220
|
| 318 |
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 0.1531654186521443,
|
| 321 |
-
"grad_norm": 42424.046875,
|
| 322 |
-
"learning_rate": 7.658270932607216e-07,
|
| 323 |
-
"loss": 2.0771,
|
| 324 |
-
"step": 225
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.15656909462219196,
|
| 328 |
-
"grad_norm": 17914.38671875,
|
| 329 |
-
"learning_rate": 7.828454731109599e-07,
|
| 330 |
-
"loss": 2.0777,
|
| 331 |
-
"step": 230
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.15997277059223963,
|
| 335 |
-
"grad_norm": 24413.658203125,
|
| 336 |
-
"learning_rate": 7.998638529611982e-07,
|
| 337 |
-
"loss": 2.0778,
|
| 338 |
-
"step": 235
|
| 339 |
-
},
|
| 340 |
{
|
| 341 |
"epoch": 0.16337644656228728,
|
| 342 |
-
"grad_norm":
|
| 343 |
-
"learning_rate":
|
| 344 |
-
"loss": 2.
|
| 345 |
"step": 240
|
| 346 |
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.16678012253233493,
|
| 349 |
-
"grad_norm": 22068.18359375,
|
| 350 |
-
"learning_rate": 8.339006126616747e-07,
|
| 351 |
-
"loss": 2.0772,
|
| 352 |
-
"step": 245
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 0.17018379850238258,
|
| 356 |
-
"grad_norm": 32458.771484375,
|
| 357 |
-
"learning_rate": 8.50918992511913e-07,
|
| 358 |
-
"loss": 2.0772,
|
| 359 |
-
"step": 250
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.17358747447243023,
|
| 363 |
-
"grad_norm": 31895.396484375,
|
| 364 |
-
"learning_rate": 8.679373723621512e-07,
|
| 365 |
-
"loss": 2.0775,
|
| 366 |
-
"step": 255
|
| 367 |
-
},
|
| 368 |
{
|
| 369 |
"epoch": 0.17699115044247787,
|
| 370 |
-
"grad_norm":
|
| 371 |
-
"learning_rate":
|
| 372 |
-
"loss": 2.
|
| 373 |
"step": 260
|
| 374 |
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.18039482641252552,
|
| 377 |
-
"grad_norm": 35021.390625,
|
| 378 |
-
"learning_rate": 9.019741320626277e-07,
|
| 379 |
-
"loss": 2.0766,
|
| 380 |
-
"step": 265
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.18379850238257317,
|
| 384 |
-
"grad_norm": 34462.171875,
|
| 385 |
-
"learning_rate": 9.189925119128659e-07,
|
| 386 |
-
"loss": 2.0764,
|
| 387 |
-
"step": 270
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.18720217835262082,
|
| 391 |
-
"grad_norm": 24412.595703125,
|
| 392 |
-
"learning_rate": 9.360108917631042e-07,
|
| 393 |
-
"loss": 2.0775,
|
| 394 |
-
"step": 275
|
| 395 |
-
},
|
| 396 |
{
|
| 397 |
"epoch": 0.1906058543226685,
|
| 398 |
-
"grad_norm":
|
| 399 |
-
"learning_rate":
|
| 400 |
-
"loss": 2.
|
| 401 |
"step": 280
|
| 402 |
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.19400953029271614,
|
| 405 |
-
"grad_norm": 32220.4921875,
|
| 406 |
-
"learning_rate": 9.700476514635807e-07,
|
| 407 |
-
"loss": 2.0757,
|
| 408 |
-
"step": 285
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.1974132062627638,
|
| 412 |
-
"grad_norm": 37819.15234375,
|
| 413 |
-
"learning_rate": 9.87066031313819e-07,
|
| 414 |
-
"loss": 2.0759,
|
| 415 |
-
"step": 290
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.20081688223281144,
|
| 419 |
-
"grad_norm": 33033.5546875,
|
| 420 |
-
"learning_rate": 1.0040844111640573e-06,
|
| 421 |
-
"loss": 2.0757,
|
| 422 |
-
"step": 295
|
| 423 |
-
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2042205582028591,
|
| 426 |
-
"grad_norm":
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 2.
|
| 429 |
"step": 300
|
| 430 |
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.20762423417290674,
|
| 433 |
-
"grad_norm": 37499.5078125,
|
| 434 |
-
"learning_rate": 1.0381211708645337e-06,
|
| 435 |
-
"loss": 2.0754,
|
| 436 |
-
"step": 305
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.21102791014295438,
|
| 440 |
-
"grad_norm": 29036.44921875,
|
| 441 |
-
"learning_rate": 1.0551395507147721e-06,
|
| 442 |
-
"loss": 2.0759,
|
| 443 |
-
"step": 310
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.21443158611300203,
|
| 447 |
-
"grad_norm": 21556.12890625,
|
| 448 |
-
"learning_rate": 1.0721579305650103e-06,
|
| 449 |
-
"loss": 2.0776,
|
| 450 |
-
"step": 315
|
| 451 |
-
},
|
| 452 |
{
|
| 453 |
"epoch": 0.21783526208304968,
|
| 454 |
-
"grad_norm":
|
| 455 |
-
"learning_rate":
|
| 456 |
-
"loss": 2.
|
| 457 |
"step": 320
|
| 458 |
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 0.22123893805309736,
|
| 461 |
-
"grad_norm": 40844.2421875,
|
| 462 |
-
"learning_rate": 1.106194690265487e-06,
|
| 463 |
-
"loss": 2.076,
|
| 464 |
-
"step": 325
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.224642614023145,
|
| 468 |
-
"grad_norm": 35128.1875,
|
| 469 |
-
"learning_rate": 1.1232130701157251e-06,
|
| 470 |
-
"loss": 2.0763,
|
| 471 |
-
"step": 330
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.22804628999319265,
|
| 475 |
-
"grad_norm": 24311.32421875,
|
| 476 |
-
"learning_rate": 1.1402314499659633e-06,
|
| 477 |
-
"loss": 2.074,
|
| 478 |
-
"step": 335
|
| 479 |
-
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2314499659632403,
|
| 482 |
-
"grad_norm":
|
| 483 |
-
"learning_rate":
|
| 484 |
-
"loss": 2.
|
| 485 |
"step": 340
|
| 486 |
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.23485364193328795,
|
| 489 |
-
"grad_norm": 24835.763671875,
|
| 490 |
-
"learning_rate": 1.17426820966644e-06,
|
| 491 |
-
"loss": 2.0756,
|
| 492 |
-
"step": 345
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.2382573179033356,
|
| 496 |
-
"grad_norm": 33911.7734375,
|
| 497 |
-
"learning_rate": 1.1912865895166781e-06,
|
| 498 |
-
"loss": 2.0735,
|
| 499 |
-
"step": 350
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.24166099387338325,
|
| 503 |
-
"grad_norm": 37201.17578125,
|
| 504 |
-
"learning_rate": 1.2083049693669163e-06,
|
| 505 |
-
"loss": 2.0737,
|
| 506 |
-
"step": 355
|
| 507 |
-
},
|
| 508 |
{
|
| 509 |
"epoch": 0.2450646698434309,
|
| 510 |
-
"grad_norm":
|
| 511 |
-
"learning_rate":
|
| 512 |
-
"loss": 2.
|
| 513 |
"step": 360
|
| 514 |
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.24846834581347857,
|
| 517 |
-
"grad_norm": 24953.693359375,
|
| 518 |
-
"learning_rate": 1.242341729067393e-06,
|
| 519 |
-
"loss": 2.0746,
|
| 520 |
-
"step": 365
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.2518720217835262,
|
| 524 |
-
"grad_norm": 35145.75390625,
|
| 525 |
-
"learning_rate": 1.2593601089176311e-06,
|
| 526 |
-
"loss": 2.0724,
|
| 527 |
-
"step": 370
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.25527569775357384,
|
| 531 |
-
"grad_norm": 40706.12109375,
|
| 532 |
-
"learning_rate": 1.2763784887678693e-06,
|
| 533 |
-
"loss": 2.0722,
|
| 534 |
-
"step": 375
|
| 535 |
-
},
|
| 536 |
{
|
| 537 |
"epoch": 0.2586793737236215,
|
| 538 |
-
"grad_norm":
|
| 539 |
-
"learning_rate":
|
| 540 |
-
"loss": 2.
|
| 541 |
"step": 380
|
| 542 |
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.2620830496936692,
|
| 545 |
-
"grad_norm": 36275.78515625,
|
| 546 |
-
"learning_rate": 1.3104152484683461e-06,
|
| 547 |
-
"loss": 2.0733,
|
| 548 |
-
"step": 385
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.26548672566371684,
|
| 552 |
-
"grad_norm": 68355.515625,
|
| 553 |
-
"learning_rate": 1.3274336283185843e-06,
|
| 554 |
-
"loss": 2.0716,
|
| 555 |
-
"step": 390
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.2688904016337645,
|
| 559 |
-
"grad_norm": 103645.6015625,
|
| 560 |
-
"learning_rate": 1.3444520081688225e-06,
|
| 561 |
-
"loss": 2.0709,
|
| 562 |
-
"step": 395
|
| 563 |
-
},
|
| 564 |
{
|
| 565 |
"epoch": 0.27229407760381213,
|
| 566 |
-
"grad_norm":
|
| 567 |
-
"learning_rate":
|
| 568 |
-
"loss": 2.
|
| 569 |
"step": 400
|
| 570 |
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.2756977535738598,
|
| 573 |
-
"grad_norm": 25283.515625,
|
| 574 |
-
"learning_rate": 1.378488767869299e-06,
|
| 575 |
-
"loss": 2.0724,
|
| 576 |
-
"step": 405
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.27910142954390743,
|
| 580 |
-
"grad_norm": 47131.55859375,
|
| 581 |
-
"learning_rate": 1.3955071477195373e-06,
|
| 582 |
-
"loss": 2.0734,
|
| 583 |
-
"step": 410
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.2825051055139551,
|
| 587 |
-
"grad_norm": 63215.078125,
|
| 588 |
-
"learning_rate": 1.4125255275697755e-06,
|
| 589 |
-
"loss": 2.0716,
|
| 590 |
-
"step": 415
|
| 591 |
-
},
|
| 592 |
{
|
| 593 |
"epoch": 0.2859087814840027,
|
| 594 |
-
"grad_norm":
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 2.
|
| 597 |
"step": 420
|
| 598 |
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.2893124574540504,
|
| 601 |
-
"grad_norm": 30855.017578125,
|
| 602 |
-
"learning_rate": 1.446562287270252e-06,
|
| 603 |
-
"loss": 2.0705,
|
| 604 |
-
"step": 425
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.292716133424098,
|
| 608 |
-
"grad_norm": 28640.982421875,
|
| 609 |
-
"learning_rate": 1.4635806671204901e-06,
|
| 610 |
-
"loss": 2.0732,
|
| 611 |
-
"step": 430
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.29611980939414567,
|
| 615 |
-
"grad_norm": 36892.05078125,
|
| 616 |
-
"learning_rate": 1.4805990469707285e-06,
|
| 617 |
-
"loss": 2.0702,
|
| 618 |
-
"step": 435
|
| 619 |
-
},
|
| 620 |
{
|
| 621 |
"epoch": 0.2995234853641933,
|
| 622 |
-
"grad_norm":
|
| 623 |
-
"learning_rate":
|
| 624 |
-
"loss": 2.
|
| 625 |
"step": 440
|
| 626 |
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.30292716133424097,
|
| 629 |
-
"grad_norm": 42311.51171875,
|
| 630 |
-
"learning_rate": 1.514635806671205e-06,
|
| 631 |
-
"loss": 2.0675,
|
| 632 |
-
"step": 445
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 0.3063308373042886,
|
| 636 |
-
"grad_norm": 66678.203125,
|
| 637 |
-
"learning_rate": 1.5316541865214431e-06,
|
| 638 |
-
"loss": 2.0666,
|
| 639 |
-
"step": 450
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.30973451327433627,
|
| 643 |
-
"grad_norm": 32896.28515625,
|
| 644 |
-
"learning_rate": 1.5486725663716816e-06,
|
| 645 |
-
"loss": 2.0669,
|
| 646 |
-
"step": 455
|
| 647 |
-
},
|
| 648 |
{
|
| 649 |
"epoch": 0.3131381892443839,
|
| 650 |
-
"grad_norm":
|
| 651 |
-
"learning_rate":
|
| 652 |
-
"loss": 2.
|
| 653 |
"step": 460
|
| 654 |
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.31654186521443156,
|
| 657 |
-
"grad_norm": 26422.58984375,
|
| 658 |
-
"learning_rate": 1.582709326072158e-06,
|
| 659 |
-
"loss": 2.0695,
|
| 660 |
-
"step": 465
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 0.31994554118447927,
|
| 664 |
-
"grad_norm": 36145.09375,
|
| 665 |
-
"learning_rate": 1.5997277059223964e-06,
|
| 666 |
-
"loss": 2.0651,
|
| 667 |
-
"step": 470
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 0.3233492171545269,
|
| 671 |
-
"grad_norm": 23801.0078125,
|
| 672 |
-
"learning_rate": 1.6167460857726348e-06,
|
| 673 |
-
"loss": 2.065,
|
| 674 |
-
"step": 475
|
| 675 |
-
},
|
| 676 |
{
|
| 677 |
"epoch": 0.32675289312457456,
|
| 678 |
-
"grad_norm":
|
| 679 |
-
"learning_rate":
|
| 680 |
-
"loss": 2.
|
| 681 |
"step": 480
|
| 682 |
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 0.3301565690946222,
|
| 685 |
-
"grad_norm": 26096.126953125,
|
| 686 |
-
"learning_rate": 1.6507828454731112e-06,
|
| 687 |
-
"loss": 2.069,
|
| 688 |
-
"step": 485
|
| 689 |
-
},
|
| 690 |
-
{
|
| 691 |
-
"epoch": 0.33356024506466986,
|
| 692 |
-
"grad_norm": 32410.6015625,
|
| 693 |
-
"learning_rate": 1.6678012253233494e-06,
|
| 694 |
-
"loss": 2.0631,
|
| 695 |
-
"step": 490
|
| 696 |
-
},
|
| 697 |
-
{
|
| 698 |
-
"epoch": 0.3369639210347175,
|
| 699 |
-
"grad_norm": 32699.9609375,
|
| 700 |
-
"learning_rate": 1.6848196051735878e-06,
|
| 701 |
-
"loss": 2.0679,
|
| 702 |
-
"step": 495
|
| 703 |
-
},
|
| 704 |
{
|
| 705 |
"epoch": 0.34036759700476515,
|
| 706 |
-
"grad_norm":
|
| 707 |
-
"learning_rate":
|
| 708 |
-
"loss": 2.
|
| 709 |
"step": 500
|
| 710 |
},
|
| 711 |
{
|
| 712 |
-
"epoch": 0.
|
| 713 |
-
"
|
| 714 |
-
"
|
| 715 |
-
"
|
| 716 |
-
"
|
| 717 |
-
|
| 718 |
-
|
| 719 |
-
"epoch": 0.34717494894486045,
|
| 720 |
-
"grad_norm": 36909.84765625,
|
| 721 |
-
"learning_rate": 1.7358747447243024e-06,
|
| 722 |
-
"loss": 2.0601,
|
| 723 |
-
"step": 510
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 0.3505786249149081,
|
| 727 |
-
"grad_norm": 29596.615234375,
|
| 728 |
-
"learning_rate": 1.7528931245745406e-06,
|
| 729 |
-
"loss": 2.0567,
|
| 730 |
-
"step": 515
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.35398230088495575,
|
| 734 |
-
"grad_norm":
|
| 735 |
-
"learning_rate":
|
| 736 |
-
"loss":
|
| 737 |
"step": 520
|
| 738 |
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 0.3573859768550034,
|
| 741 |
-
"grad_norm": 20688.9375,
|
| 742 |
-
"learning_rate": 1.7869298842750172e-06,
|
| 743 |
-
"loss": 2.0644,
|
| 744 |
-
"step": 525
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 0.36078965282505104,
|
| 748 |
-
"grad_norm": 43875.91015625,
|
| 749 |
-
"learning_rate": 1.8039482641252554e-06,
|
| 750 |
-
"loss": 2.0583,
|
| 751 |
-
"step": 530
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.3641933287950987,
|
| 755 |
-
"grad_norm": 43107.44921875,
|
| 756 |
-
"learning_rate": 1.8209666439754936e-06,
|
| 757 |
-
"loss": 2.0611,
|
| 758 |
-
"step": 535
|
| 759 |
-
},
|
| 760 |
{
|
| 761 |
"epoch": 0.36759700476514634,
|
| 762 |
-
"grad_norm":
|
| 763 |
-
"learning_rate":
|
| 764 |
-
"loss":
|
| 765 |
"step": 540
|
| 766 |
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 0.371000680735194,
|
| 769 |
-
"grad_norm": 25677.87890625,
|
| 770 |
-
"learning_rate": 1.8550034036759702e-06,
|
| 771 |
-
"loss": 2.0581,
|
| 772 |
-
"step": 545
|
| 773 |
-
},
|
| 774 |
-
{
|
| 775 |
-
"epoch": 0.37440435670524164,
|
| 776 |
-
"grad_norm": 32200.326171875,
|
| 777 |
-
"learning_rate": 1.8720217835262084e-06,
|
| 778 |
-
"loss": 2.0529,
|
| 779 |
-
"step": 550
|
| 780 |
-
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 0.3778080326752893,
|
| 783 |
-
"grad_norm": 59106.46484375,
|
| 784 |
-
"learning_rate": 1.8890401633764466e-06,
|
| 785 |
-
"loss": 2.0538,
|
| 786 |
-
"step": 555
|
| 787 |
-
},
|
| 788 |
{
|
| 789 |
"epoch": 0.381211708645337,
|
| 790 |
-
"grad_norm":
|
| 791 |
-
"learning_rate":
|
| 792 |
-
"loss":
|
| 793 |
"step": 560
|
| 794 |
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 0.38461538461538464,
|
| 797 |
-
"grad_norm": 33576.88671875,
|
| 798 |
-
"learning_rate": 1.9230769230769234e-06,
|
| 799 |
-
"loss": 2.0506,
|
| 800 |
-
"step": 565
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 0.3880190605854323,
|
| 804 |
-
"grad_norm": 50821.08984375,
|
| 805 |
-
"learning_rate": 1.9400953029271614e-06,
|
| 806 |
-
"loss": 2.0547,
|
| 807 |
-
"step": 570
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 0.39142273655547993,
|
| 811 |
-
"grad_norm": 29727.3125,
|
| 812 |
-
"learning_rate": 1.9571136827774e-06,
|
| 813 |
-
"loss": 2.0549,
|
| 814 |
-
"step": 575
|
| 815 |
-
},
|
| 816 |
{
|
| 817 |
"epoch": 0.3948264125255276,
|
| 818 |
-
"grad_norm":
|
| 819 |
-
"learning_rate":
|
| 820 |
-
"loss":
|
| 821 |
"step": 580
|
| 822 |
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 0.39823008849557523,
|
| 825 |
-
"grad_norm": 43372.19921875,
|
| 826 |
-
"learning_rate": 1.991150442477876e-06,
|
| 827 |
-
"loss": 2.0459,
|
| 828 |
-
"step": 585
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 0.4016337644656229,
|
| 832 |
-
"grad_norm": 35458.55859375,
|
| 833 |
-
"learning_rate": 2.0081688223281146e-06,
|
| 834 |
-
"loss": 2.0527,
|
| 835 |
-
"step": 590
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 0.4050374404356705,
|
| 839 |
-
"grad_norm": 38893.390625,
|
| 840 |
-
"learning_rate": 2.0251872021783526e-06,
|
| 841 |
-
"loss": 2.0507,
|
| 842 |
-
"step": 595
|
| 843 |
-
},
|
| 844 |
{
|
| 845 |
"epoch": 0.4084411164057182,
|
| 846 |
-
"grad_norm":
|
| 847 |
-
"learning_rate":
|
| 848 |
-
"loss":
|
| 849 |
"step": 600
|
| 850 |
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 0.4118447923757658,
|
| 853 |
-
"grad_norm": 37080.59765625,
|
| 854 |
-
"learning_rate": 2.0592239618788294e-06,
|
| 855 |
-
"loss": 2.0467,
|
| 856 |
-
"step": 605
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 0.41524846834581347,
|
| 860 |
-
"grad_norm": 38622.73046875,
|
| 861 |
-
"learning_rate": 2.0762423417290674e-06,
|
| 862 |
-
"loss": 2.0429,
|
| 863 |
-
"step": 610
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 0.4186521443158611,
|
| 867 |
-
"grad_norm": 27659.953125,
|
| 868 |
-
"learning_rate": 2.093260721579306e-06,
|
| 869 |
-
"loss": 2.0407,
|
| 870 |
-
"step": 615
|
| 871 |
-
},
|
| 872 |
{
|
| 873 |
"epoch": 0.42205582028590877,
|
| 874 |
-
"grad_norm":
|
| 875 |
-
"learning_rate":
|
| 876 |
-
"loss":
|
| 877 |
"step": 620
|
| 878 |
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 0.4254594962559564,
|
| 881 |
-
"grad_norm": 23164.375,
|
| 882 |
-
"learning_rate": 2.127297481279782e-06,
|
| 883 |
-
"loss": 2.044,
|
| 884 |
-
"step": 625
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 0.42886317222600406,
|
| 888 |
-
"grad_norm": 24133.185546875,
|
| 889 |
-
"learning_rate": 2.1443158611300206e-06,
|
| 890 |
-
"loss": 2.0353,
|
| 891 |
-
"step": 630
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 0.4322668481960517,
|
| 895 |
-
"grad_norm": 130830.8125,
|
| 896 |
-
"learning_rate": 2.1613342409802586e-06,
|
| 897 |
-
"loss": 2.032,
|
| 898 |
-
"step": 635
|
| 899 |
-
},
|
| 900 |
{
|
| 901 |
"epoch": 0.43567052416609936,
|
| 902 |
-
"grad_norm":
|
| 903 |
-
"learning_rate":
|
| 904 |
-
"loss":
|
| 905 |
"step": 640
|
| 906 |
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 0.43907420013614706,
|
| 909 |
-
"grad_norm": 44940.66796875,
|
| 910 |
-
"learning_rate": 2.1953710006807354e-06,
|
| 911 |
-
"loss": 2.0288,
|
| 912 |
-
"step": 645
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 0.4424778761061947,
|
| 916 |
-
"grad_norm": 51611.99609375,
|
| 917 |
-
"learning_rate": 2.212389380530974e-06,
|
| 918 |
-
"loss": 2.0238,
|
| 919 |
-
"step": 650
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 0.44588155207624236,
|
| 923 |
-
"grad_norm": 45395.26953125,
|
| 924 |
-
"learning_rate": 2.229407760381212e-06,
|
| 925 |
-
"loss": 2.0268,
|
| 926 |
-
"step": 655
|
| 927 |
-
},
|
| 928 |
{
|
| 929 |
"epoch": 0.44928522804629,
|
| 930 |
-
"grad_norm":
|
| 931 |
-
"learning_rate":
|
| 932 |
-
"loss":
|
| 933 |
"step": 660
|
| 934 |
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 0.45268890401633766,
|
| 937 |
-
"grad_norm": 34002.0078125,
|
| 938 |
-
"learning_rate": 2.2634445200816886e-06,
|
| 939 |
-
"loss": 2.018,
|
| 940 |
-
"step": 665
|
| 941 |
-
},
|
| 942 |
-
{
|
| 943 |
-
"epoch": 0.4560925799863853,
|
| 944 |
-
"grad_norm": 47849.36328125,
|
| 945 |
-
"learning_rate": 2.2804628999319266e-06,
|
| 946 |
-
"loss": 2.0141,
|
| 947 |
-
"step": 670
|
| 948 |
-
},
|
| 949 |
-
{
|
| 950 |
-
"epoch": 0.45949625595643295,
|
| 951 |
-
"grad_norm": 37371.65234375,
|
| 952 |
-
"learning_rate": 2.297481279782165e-06,
|
| 953 |
-
"loss": 2.0229,
|
| 954 |
-
"step": 675
|
| 955 |
-
},
|
| 956 |
{
|
| 957 |
"epoch": 0.4628999319264806,
|
| 958 |
-
"grad_norm":
|
| 959 |
-
"learning_rate":
|
| 960 |
-
"loss":
|
| 961 |
"step": 680
|
| 962 |
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 0.46630360789652825,
|
| 965 |
-
"grad_norm": 74534.9296875,
|
| 966 |
-
"learning_rate": 2.3315180394826414e-06,
|
| 967 |
-
"loss": 2.0203,
|
| 968 |
-
"step": 685
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 0.4697072838665759,
|
| 972 |
-
"grad_norm": 53662.75390625,
|
| 973 |
-
"learning_rate": 2.34853641933288e-06,
|
| 974 |
-
"loss": 2.0565,
|
| 975 |
-
"step": 690
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 0.47311095983662355,
|
| 979 |
-
"grad_norm": 56688.16796875,
|
| 980 |
-
"learning_rate": 2.365554799183118e-06,
|
| 981 |
-
"loss": 2.029,
|
| 982 |
-
"step": 695
|
| 983 |
-
},
|
| 984 |
{
|
| 985 |
"epoch": 0.4765146358066712,
|
| 986 |
-
"grad_norm":
|
| 987 |
-
"learning_rate":
|
| 988 |
-
"loss":
|
| 989 |
"step": 700
|
| 990 |
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 0.47991831177671884,
|
| 993 |
-
"grad_norm": 75029.7890625,
|
| 994 |
-
"learning_rate": 2.3995915588835942e-06,
|
| 995 |
-
"loss": 2.0314,
|
| 996 |
-
"step": 705
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 0.4833219877467665,
|
| 1000 |
-
"grad_norm": 62975.13671875,
|
| 1001 |
-
"learning_rate": 2.4166099387338326e-06,
|
| 1002 |
-
"loss": 2.0074,
|
| 1003 |
-
"step": 710
|
| 1004 |
-
},
|
| 1005 |
-
{
|
| 1006 |
-
"epoch": 0.48672566371681414,
|
| 1007 |
-
"grad_norm": 47454.12109375,
|
| 1008 |
-
"learning_rate": 2.433628318584071e-06,
|
| 1009 |
-
"loss": 2.0258,
|
| 1010 |
-
"step": 715
|
| 1011 |
-
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.4901293396868618,
|
| 1014 |
-
"grad_norm":
|
| 1015 |
-
"learning_rate":
|
| 1016 |
-
"loss":
|
| 1017 |
"step": 720
|
| 1018 |
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 0.49353301565690944,
|
| 1021 |
-
"grad_norm": 124756.7109375,
|
| 1022 |
-
"learning_rate": 2.4676650782845474e-06,
|
| 1023 |
-
"loss": 2.0109,
|
| 1024 |
-
"step": 725
|
| 1025 |
-
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 0.49693669162695714,
|
| 1028 |
-
"grad_norm": 35529.78515625,
|
| 1029 |
-
"learning_rate": 2.484683458134786e-06,
|
| 1030 |
-
"loss": 2.0225,
|
| 1031 |
-
"step": 730
|
| 1032 |
-
},
|
| 1033 |
-
{
|
| 1034 |
-
"epoch": 0.5003403675970047,
|
| 1035 |
-
"grad_norm": 69232.0078125,
|
| 1036 |
-
"learning_rate": 2.501701837985024e-06,
|
| 1037 |
-
"loss": 2.0049,
|
| 1038 |
-
"step": 735
|
| 1039 |
-
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.5037440435670524,
|
| 1042 |
-
"grad_norm":
|
| 1043 |
-
"learning_rate":
|
| 1044 |
-
"loss":
|
| 1045 |
"step": 740
|
| 1046 |
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 0.5071477195371,
|
| 1049 |
-
"grad_norm": 34156.91796875,
|
| 1050 |
-
"learning_rate": 2.5357385976855002e-06,
|
| 1051 |
-
"loss": 2.0114,
|
| 1052 |
-
"step": 745
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 0.5105513955071477,
|
| 1056 |
-
"grad_norm": 66655.0,
|
| 1057 |
-
"learning_rate": 2.5527569775357386e-06,
|
| 1058 |
-
"loss": 2.0115,
|
| 1059 |
-
"step": 750
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 0.5139550714771953,
|
| 1063 |
-
"grad_norm": 87704.65625,
|
| 1064 |
-
"learning_rate": 2.5697753573859766e-06,
|
| 1065 |
-
"loss": 1.9765,
|
| 1066 |
-
"step": 755
|
| 1067 |
-
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.517358747447243,
|
| 1070 |
-
"grad_norm":
|
| 1071 |
-
"learning_rate":
|
| 1072 |
-
"loss": 1.
|
| 1073 |
"step": 760
|
| 1074 |
},
|
| 1075 |
-
{
|
| 1076 |
-
"epoch": 0.5207624234172906,
|
| 1077 |
-
"grad_norm": 54837.47265625,
|
| 1078 |
-
"learning_rate": 2.6038121170864534e-06,
|
| 1079 |
-
"loss": 1.9832,
|
| 1080 |
-
"step": 765
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 0.5241660993873384,
|
| 1084 |
-
"grad_norm": 72864.1640625,
|
| 1085 |
-
"learning_rate": 2.6208304969366923e-06,
|
| 1086 |
-
"loss": 2.0094,
|
| 1087 |
-
"step": 770
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 0.527569775357386,
|
| 1091 |
-
"grad_norm": 49820.19140625,
|
| 1092 |
-
"learning_rate": 2.6378488767869303e-06,
|
| 1093 |
-
"loss": 1.9846,
|
| 1094 |
-
"step": 775
|
| 1095 |
-
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.5309734513274337,
|
| 1098 |
-
"grad_norm":
|
| 1099 |
-
"learning_rate":
|
| 1100 |
-
"loss": 1.
|
| 1101 |
"step": 780
|
| 1102 |
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 0.5343771272974813,
|
| 1105 |
-
"grad_norm": 51874.91015625,
|
| 1106 |
-
"learning_rate": 2.6718856364874067e-06,
|
| 1107 |
-
"loss": 2.0307,
|
| 1108 |
-
"step": 785
|
| 1109 |
-
},
|
| 1110 |
-
{
|
| 1111 |
-
"epoch": 0.537780803267529,
|
| 1112 |
-
"grad_norm": 36787.6015625,
|
| 1113 |
-
"learning_rate": 2.688904016337645e-06,
|
| 1114 |
-
"loss": 1.9709,
|
| 1115 |
-
"step": 790
|
| 1116 |
-
},
|
| 1117 |
-
{
|
| 1118 |
-
"epoch": 0.5411844792375766,
|
| 1119 |
-
"grad_norm": 61949.953125,
|
| 1120 |
-
"learning_rate": 2.7059223961878835e-06,
|
| 1121 |
-
"loss": 1.9839,
|
| 1122 |
-
"step": 795
|
| 1123 |
-
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.5445881552076243,
|
| 1126 |
-
"grad_norm":
|
| 1127 |
-
"learning_rate":
|
| 1128 |
-
"loss": 1.
|
| 1129 |
"step": 800
|
| 1130 |
},
|
| 1131 |
-
{
|
| 1132 |
-
"epoch": 0.5479918311776719,
|
| 1133 |
-
"grad_norm": 58997.12890625,
|
| 1134 |
-
"learning_rate": 2.73995915588836e-06,
|
| 1135 |
-
"loss": 2.0024,
|
| 1136 |
-
"step": 805
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 0.5513955071477196,
|
| 1140 |
-
"grad_norm": 75211.9765625,
|
| 1141 |
-
"learning_rate": 2.756977535738598e-06,
|
| 1142 |
-
"loss": 1.9998,
|
| 1143 |
-
"step": 810
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 0.5547991831177672,
|
| 1147 |
-
"grad_norm": 57823.1640625,
|
| 1148 |
-
"learning_rate": 2.7739959155888363e-06,
|
| 1149 |
-
"loss": 1.978,
|
| 1150 |
-
"step": 815
|
| 1151 |
-
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.5582028590878149,
|
| 1154 |
-
"grad_norm":
|
| 1155 |
-
"learning_rate":
|
| 1156 |
-
"loss": 1.
|
| 1157 |
"step": 820
|
| 1158 |
},
|
| 1159 |
-
{
|
| 1160 |
-
"epoch": 0.5616065350578625,
|
| 1161 |
-
"grad_norm": 61511.76953125,
|
| 1162 |
-
"learning_rate": 2.8080326752893127e-06,
|
| 1163 |
-
"loss": 1.9938,
|
| 1164 |
-
"step": 825
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 0.5650102110279102,
|
| 1168 |
-
"grad_norm": 56712.01171875,
|
| 1169 |
-
"learning_rate": 2.825051055139551e-06,
|
| 1170 |
-
"loss": 1.975,
|
| 1171 |
-
"step": 830
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 0.5684138869979578,
|
| 1175 |
-
"grad_norm": 85587.9765625,
|
| 1176 |
-
"learning_rate": 2.842069434989789e-06,
|
| 1177 |
-
"loss": 1.9637,
|
| 1178 |
-
"step": 835
|
| 1179 |
-
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.5718175629680055,
|
| 1182 |
-
"grad_norm":
|
| 1183 |
-
"learning_rate":
|
| 1184 |
-
"loss": 1.
|
| 1185 |
"step": 840
|
| 1186 |
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 0.5752212389380531,
|
| 1189 |
-
"grad_norm": 56330.55859375,
|
| 1190 |
-
"learning_rate": 2.876106194690266e-06,
|
| 1191 |
-
"loss": 1.9586,
|
| 1192 |
-
"step": 845
|
| 1193 |
-
},
|
| 1194 |
-
{
|
| 1195 |
-
"epoch": 0.5786249149081008,
|
| 1196 |
-
"grad_norm": 60672.6171875,
|
| 1197 |
-
"learning_rate": 2.893124574540504e-06,
|
| 1198 |
-
"loss": 1.976,
|
| 1199 |
-
"step": 850
|
| 1200 |
-
},
|
| 1201 |
-
{
|
| 1202 |
-
"epoch": 0.5820285908781484,
|
| 1203 |
-
"grad_norm": 50871.8359375,
|
| 1204 |
-
"learning_rate": 2.9101429543907423e-06,
|
| 1205 |
-
"loss": 1.9474,
|
| 1206 |
-
"step": 855
|
| 1207 |
-
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.585432266848196,
|
| 1210 |
-
"grad_norm":
|
| 1211 |
-
"learning_rate":
|
| 1212 |
-
"loss": 1.
|
| 1213 |
"step": 860
|
| 1214 |
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 0.5888359428182437,
|
| 1217 |
-
"grad_norm": 45848.640625,
|
| 1218 |
-
"learning_rate": 2.9441797140912187e-06,
|
| 1219 |
-
"loss": 1.9854,
|
| 1220 |
-
"step": 865
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 0.5922396187882913,
|
| 1224 |
-
"grad_norm": 76285.7109375,
|
| 1225 |
-
"learning_rate": 2.961198093941457e-06,
|
| 1226 |
-
"loss": 1.9974,
|
| 1227 |
-
"step": 870
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 0.595643294758339,
|
| 1231 |
-
"grad_norm": 71599.1796875,
|
| 1232 |
-
"learning_rate": 2.978216473791695e-06,
|
| 1233 |
-
"loss": 1.9714,
|
| 1234 |
-
"step": 875
|
| 1235 |
-
},
|
| 1236 |
{
|
| 1237 |
"epoch": 0.5990469707283866,
|
| 1238 |
-
"grad_norm":
|
| 1239 |
-
"learning_rate":
|
| 1240 |
-
"loss": 1.
|
| 1241 |
"step": 880
|
| 1242 |
},
|
| 1243 |
-
{
|
| 1244 |
-
"epoch": 0.6024506466984343,
|
| 1245 |
-
"grad_norm": 89773.8046875,
|
| 1246 |
-
"learning_rate": 3.012253233492172e-06,
|
| 1247 |
-
"loss": 1.9542,
|
| 1248 |
-
"step": 885
|
| 1249 |
-
},
|
| 1250 |
-
{
|
| 1251 |
-
"epoch": 0.6058543226684819,
|
| 1252 |
-
"grad_norm": 75365.0078125,
|
| 1253 |
-
"learning_rate": 3.02927161334241e-06,
|
| 1254 |
-
"loss": 1.9763,
|
| 1255 |
-
"step": 890
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 0.6092579986385296,
|
| 1259 |
-
"grad_norm": 104382.90625,
|
| 1260 |
-
"learning_rate": 3.0462899931926483e-06,
|
| 1261 |
-
"loss": 1.9427,
|
| 1262 |
-
"step": 895
|
| 1263 |
-
},
|
| 1264 |
{
|
| 1265 |
"epoch": 0.6126616746085772,
|
| 1266 |
-
"grad_norm":
|
| 1267 |
-
"learning_rate":
|
| 1268 |
-
"loss": 1.
|
| 1269 |
"step": 900
|
| 1270 |
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 0.6160653505786249,
|
| 1273 |
-
"grad_norm": 120366.875,
|
| 1274 |
-
"learning_rate": 3.0803267528931247e-06,
|
| 1275 |
-
"loss": 1.9407,
|
| 1276 |
-
"step": 905
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 0.6194690265486725,
|
| 1280 |
-
"grad_norm": 57796.1328125,
|
| 1281 |
-
"learning_rate": 3.097345132743363e-06,
|
| 1282 |
-
"loss": 1.9107,
|
| 1283 |
-
"step": 910
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 0.6228727025187202,
|
| 1287 |
-
"grad_norm": 81527.2265625,
|
| 1288 |
-
"learning_rate": 3.114363512593601e-06,
|
| 1289 |
-
"loss": 1.8782,
|
| 1290 |
-
"step": 915
|
| 1291 |
-
},
|
| 1292 |
{
|
| 1293 |
"epoch": 0.6262763784887678,
|
| 1294 |
-
"grad_norm":
|
| 1295 |
-
"learning_rate":
|
| 1296 |
-
"loss": 1.
|
| 1297 |
"step": 920
|
| 1298 |
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 0.6296800544588155,
|
| 1301 |
-
"grad_norm": 109943.0234375,
|
| 1302 |
-
"learning_rate": 3.1484002722940775e-06,
|
| 1303 |
-
"loss": 1.9624,
|
| 1304 |
-
"step": 925
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 0.6330837304288631,
|
| 1308 |
-
"grad_norm": 67497.53125,
|
| 1309 |
-
"learning_rate": 3.165418652144316e-06,
|
| 1310 |
-
"loss": 1.9272,
|
| 1311 |
-
"step": 930
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 0.6364874063989108,
|
| 1315 |
-
"grad_norm": 99553.6640625,
|
| 1316 |
-
"learning_rate": 3.1824370319945543e-06,
|
| 1317 |
-
"loss": 1.9557,
|
| 1318 |
-
"step": 935
|
| 1319 |
-
},
|
| 1320 |
{
|
| 1321 |
"epoch": 0.6398910823689585,
|
| 1322 |
-
"grad_norm":
|
| 1323 |
-
"learning_rate":
|
| 1324 |
-
"loss": 1.
|
| 1325 |
"step": 940
|
| 1326 |
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 0.6432947583390062,
|
| 1329 |
-
"grad_norm": 92217.9609375,
|
| 1330 |
-
"learning_rate": 3.216473791695031e-06,
|
| 1331 |
-
"loss": 1.892,
|
| 1332 |
-
"step": 945
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 0.6466984343090538,
|
| 1336 |
-
"grad_norm": 70931.03125,
|
| 1337 |
-
"learning_rate": 3.2334921715452695e-06,
|
| 1338 |
-
"loss": 1.8809,
|
| 1339 |
-
"step": 950
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 0.6501021102791015,
|
| 1343 |
-
"grad_norm": 43681.0,
|
| 1344 |
-
"learning_rate": 3.2505105513955075e-06,
|
| 1345 |
-
"loss": 1.9718,
|
| 1346 |
-
"step": 955
|
| 1347 |
-
},
|
| 1348 |
{
|
| 1349 |
"epoch": 0.6535057862491491,
|
| 1350 |
-
"grad_norm":
|
| 1351 |
-
"learning_rate":
|
| 1352 |
-
"loss": 1.
|
| 1353 |
"step": 960
|
| 1354 |
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 0.6569094622191968,
|
| 1357 |
-
"grad_norm": 60043.171875,
|
| 1358 |
-
"learning_rate": 3.284547311095984e-06,
|
| 1359 |
-
"loss": 1.931,
|
| 1360 |
-
"step": 965
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 0.6603131381892444,
|
| 1364 |
-
"grad_norm": 120798.7265625,
|
| 1365 |
-
"learning_rate": 3.3015656909462223e-06,
|
| 1366 |
-
"loss": 1.882,
|
| 1367 |
-
"step": 970
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 0.6637168141592921,
|
| 1371 |
-
"grad_norm": 77168.4453125,
|
| 1372 |
-
"learning_rate": 3.3185840707964607e-06,
|
| 1373 |
-
"loss": 1.9859,
|
| 1374 |
-
"step": 975
|
| 1375 |
-
},
|
| 1376 |
{
|
| 1377 |
"epoch": 0.6671204901293397,
|
| 1378 |
-
"grad_norm":
|
| 1379 |
-
"learning_rate":
|
| 1380 |
-
"loss": 1.
|
| 1381 |
"step": 980
|
| 1382 |
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 0.6705241660993874,
|
| 1385 |
-
"grad_norm": 140839.265625,
|
| 1386 |
-
"learning_rate": 3.352620830496937e-06,
|
| 1387 |
-
"loss": 1.8859,
|
| 1388 |
-
"step": 985
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 0.673927842069435,
|
| 1392 |
-
"grad_norm": 62714.12109375,
|
| 1393 |
-
"learning_rate": 3.3696392103471755e-06,
|
| 1394 |
-
"loss": 1.9197,
|
| 1395 |
-
"step": 990
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 0.6773315180394827,
|
| 1399 |
-
"grad_norm": 125514.3359375,
|
| 1400 |
-
"learning_rate": 3.3866575901974135e-06,
|
| 1401 |
-
"loss": 1.9035,
|
| 1402 |
-
"step": 995
|
| 1403 |
-
},
|
| 1404 |
{
|
| 1405 |
"epoch": 0.6807351940095303,
|
| 1406 |
-
"grad_norm":
|
| 1407 |
-
"learning_rate":
|
| 1408 |
-
"loss": 1.
|
| 1409 |
"step": 1000
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
-
"epoch": 0.
|
| 1413 |
-
"
|
| 1414 |
-
"
|
| 1415 |
-
"
|
| 1416 |
-
"
|
| 1417 |
-
|
| 1418 |
-
|
| 1419 |
-
"epoch": 0.6875425459496256,
|
| 1420 |
-
"grad_norm": 97830.4453125,
|
| 1421 |
-
"learning_rate": 3.4377127297481283e-06,
|
| 1422 |
-
"loss": 1.9096,
|
| 1423 |
-
"step": 1010
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 0.6909462219196733,
|
| 1427 |
-
"grad_norm": 112141.9140625,
|
| 1428 |
-
"learning_rate": 3.4547311095983668e-06,
|
| 1429 |
-
"loss": 1.9033,
|
| 1430 |
-
"step": 1015
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 0.6943498978897209,
|
| 1434 |
-
"grad_norm":
|
| 1435 |
-
"learning_rate":
|
| 1436 |
-
"loss": 1.
|
| 1437 |
"step": 1020
|
| 1438 |
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 0.6977535738597685,
|
| 1441 |
-
"grad_norm": 109543.625,
|
| 1442 |
-
"learning_rate": 3.488767869298843e-06,
|
| 1443 |
-
"loss": 1.8608,
|
| 1444 |
-
"step": 1025
|
| 1445 |
-
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 0.7011572498298162,
|
| 1448 |
-
"grad_norm": 58480.5625,
|
| 1449 |
-
"learning_rate": 3.505786249149081e-06,
|
| 1450 |
-
"loss": 1.8792,
|
| 1451 |
-
"step": 1030
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 0.7045609257998638,
|
| 1455 |
-
"grad_norm": 110749.9296875,
|
| 1456 |
-
"learning_rate": 3.5228046289993195e-06,
|
| 1457 |
-
"loss": 1.8449,
|
| 1458 |
-
"step": 1035
|
| 1459 |
-
},
|
| 1460 |
{
|
| 1461 |
"epoch": 0.7079646017699115,
|
| 1462 |
-
"grad_norm":
|
| 1463 |
-
"learning_rate":
|
| 1464 |
-
"loss": 1.
|
| 1465 |
"step": 1040
|
| 1466 |
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 0.7113682777399591,
|
| 1469 |
-
"grad_norm": 90031.5546875,
|
| 1470 |
-
"learning_rate": 3.556841388699796e-06,
|
| 1471 |
-
"loss": 1.858,
|
| 1472 |
-
"step": 1045
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.7147719537100068,
|
| 1476 |
-
"grad_norm": 133713.484375,
|
| 1477 |
-
"learning_rate": 3.5738597685500344e-06,
|
| 1478 |
-
"loss": 1.92,
|
| 1479 |
-
"step": 1050
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.7181756296800544,
|
| 1483 |
-
"grad_norm": 120697.03125,
|
| 1484 |
-
"learning_rate": 3.5908781484002723e-06,
|
| 1485 |
-
"loss": 1.9957,
|
| 1486 |
-
"step": 1055
|
| 1487 |
-
},
|
| 1488 |
{
|
| 1489 |
"epoch": 0.7215793056501021,
|
| 1490 |
-
"grad_norm":
|
| 1491 |
-
"learning_rate":
|
| 1492 |
-
"loss": 1.
|
| 1493 |
"step": 1060
|
| 1494 |
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 0.7249829816201497,
|
| 1497 |
-
"grad_norm": 100709.84375,
|
| 1498 |
-
"learning_rate": 3.624914908100749e-06,
|
| 1499 |
-
"loss": 1.8674,
|
| 1500 |
-
"step": 1065
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 0.7283866575901974,
|
| 1504 |
-
"grad_norm": 145894.46875,
|
| 1505 |
-
"learning_rate": 3.641933287950987e-06,
|
| 1506 |
-
"loss": 1.8018,
|
| 1507 |
-
"step": 1070
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 0.731790333560245,
|
| 1511 |
-
"grad_norm": 107045.4609375,
|
| 1512 |
-
"learning_rate": 3.6589516678012256e-06,
|
| 1513 |
-
"loss": 1.8992,
|
| 1514 |
-
"step": 1075
|
| 1515 |
-
},
|
| 1516 |
{
|
| 1517 |
"epoch": 0.7351940095302927,
|
| 1518 |
-
"grad_norm":
|
| 1519 |
-
"learning_rate":
|
| 1520 |
-
"loss": 1.
|
| 1521 |
"step": 1080
|
| 1522 |
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 0.7385976855003403,
|
| 1525 |
-
"grad_norm": 66000.2578125,
|
| 1526 |
-
"learning_rate": 3.692988427501702e-06,
|
| 1527 |
-
"loss": 1.8696,
|
| 1528 |
-
"step": 1085
|
| 1529 |
-
},
|
| 1530 |
-
{
|
| 1531 |
-
"epoch": 0.742001361470388,
|
| 1532 |
-
"grad_norm": 102473.03125,
|
| 1533 |
-
"learning_rate": 3.7100068073519404e-06,
|
| 1534 |
-
"loss": 1.8025,
|
| 1535 |
-
"step": 1090
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 0.7454050374404356,
|
| 1539 |
-
"grad_norm": 58545.05859375,
|
| 1540 |
-
"learning_rate": 3.7270251872021783e-06,
|
| 1541 |
-
"loss": 1.7841,
|
| 1542 |
-
"step": 1095
|
| 1543 |
-
},
|
| 1544 |
{
|
| 1545 |
"epoch": 0.7488087134104833,
|
| 1546 |
-
"grad_norm":
|
| 1547 |
-
"learning_rate":
|
| 1548 |
-
"loss": 1.
|
| 1549 |
"step": 1100
|
| 1550 |
},
|
| 1551 |
-
{
|
| 1552 |
-
"epoch": 0.7522123893805309,
|
| 1553 |
-
"grad_norm": 165545.234375,
|
| 1554 |
-
"learning_rate": 3.7610619469026547e-06,
|
| 1555 |
-
"loss": 1.7984,
|
| 1556 |
-
"step": 1105
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 0.7556160653505786,
|
| 1560 |
-
"grad_norm": 57886.5078125,
|
| 1561 |
-
"learning_rate": 3.778080326752893e-06,
|
| 1562 |
-
"loss": 1.8817,
|
| 1563 |
-
"step": 1110
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 0.7590197413206263,
|
| 1567 |
-
"grad_norm": 135867.59375,
|
| 1568 |
-
"learning_rate": 3.795098706603132e-06,
|
| 1569 |
-
"loss": 1.9104,
|
| 1570 |
-
"step": 1115
|
| 1571 |
-
},
|
| 1572 |
{
|
| 1573 |
"epoch": 0.762423417290674,
|
| 1574 |
-
"grad_norm":
|
| 1575 |
-
"learning_rate":
|
| 1576 |
-
"loss": 1.
|
| 1577 |
"step": 1120
|
| 1578 |
},
|
| 1579 |
-
{
|
| 1580 |
-
"epoch": 0.7658270932607216,
|
| 1581 |
-
"grad_norm": 156170.046875,
|
| 1582 |
-
"learning_rate": 3.829135466303609e-06,
|
| 1583 |
-
"loss": 1.7933,
|
| 1584 |
-
"step": 1125
|
| 1585 |
-
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 0.7692307692307693,
|
| 1588 |
-
"grad_norm": 80088.84375,
|
| 1589 |
-
"learning_rate": 3.846153846153847e-06,
|
| 1590 |
-
"loss": 1.9251,
|
| 1591 |
-
"step": 1130
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 0.7726344452008169,
|
| 1595 |
-
"grad_norm": 158380.4375,
|
| 1596 |
-
"learning_rate": 3.863172226004085e-06,
|
| 1597 |
-
"loss": 1.8254,
|
| 1598 |
-
"step": 1135
|
| 1599 |
-
},
|
| 1600 |
{
|
| 1601 |
"epoch": 0.7760381211708646,
|
| 1602 |
-
"grad_norm":
|
| 1603 |
-
"learning_rate":
|
| 1604 |
-
"loss": 1.
|
| 1605 |
"step": 1140
|
| 1606 |
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 0.7794417971409122,
|
| 1609 |
-
"grad_norm": 200350.171875,
|
| 1610 |
-
"learning_rate": 3.897208985704562e-06,
|
| 1611 |
-
"loss": 1.7513,
|
| 1612 |
-
"step": 1145
|
| 1613 |
-
},
|
| 1614 |
-
{
|
| 1615 |
-
"epoch": 0.7828454731109599,
|
| 1616 |
-
"grad_norm": 145619.609375,
|
| 1617 |
-
"learning_rate": 3.9142273655548e-06,
|
| 1618 |
-
"loss": 1.7236,
|
| 1619 |
-
"step": 1150
|
| 1620 |
-
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 0.7862491490810075,
|
| 1623 |
-
"grad_norm": 107629.296875,
|
| 1624 |
-
"learning_rate": 3.9312457454050376e-06,
|
| 1625 |
-
"loss": 1.7952,
|
| 1626 |
-
"step": 1155
|
| 1627 |
-
},
|
| 1628 |
{
|
| 1629 |
"epoch": 0.7896528250510552,
|
| 1630 |
-
"grad_norm":
|
| 1631 |
-
"learning_rate":
|
| 1632 |
-
"loss": 1.
|
| 1633 |
"step": 1160
|
| 1634 |
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 0.7930565010211028,
|
| 1637 |
-
"grad_norm": 152800.21875,
|
| 1638 |
-
"learning_rate": 3.965282505105514e-06,
|
| 1639 |
-
"loss": 1.8354,
|
| 1640 |
-
"step": 1165
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 0.7964601769911505,
|
| 1644 |
-
"grad_norm": 120458.5859375,
|
| 1645 |
-
"learning_rate": 3.982300884955752e-06,
|
| 1646 |
-
"loss": 1.8895,
|
| 1647 |
-
"step": 1170
|
| 1648 |
-
},
|
| 1649 |
-
{
|
| 1650 |
-
"epoch": 0.7998638529611981,
|
| 1651 |
-
"grad_norm": 135065.109375,
|
| 1652 |
-
"learning_rate": 3.999319264805991e-06,
|
| 1653 |
-
"loss": 1.6959,
|
| 1654 |
-
"step": 1175
|
| 1655 |
-
},
|
| 1656 |
{
|
| 1657 |
"epoch": 0.8032675289312458,
|
| 1658 |
-
"grad_norm":
|
| 1659 |
-
"learning_rate": 4.
|
| 1660 |
-
"loss": 1.
|
| 1661 |
"step": 1180
|
| 1662 |
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 0.8066712049012934,
|
| 1665 |
-
"grad_norm": 149330.9375,
|
| 1666 |
-
"learning_rate": 4.033356024506467e-06,
|
| 1667 |
-
"loss": 1.7966,
|
| 1668 |
-
"step": 1185
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 0.810074880871341,
|
| 1672 |
-
"grad_norm": 162711.078125,
|
| 1673 |
-
"learning_rate": 4.050374404356705e-06,
|
| 1674 |
-
"loss": 1.7514,
|
| 1675 |
-
"step": 1190
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 0.8134785568413887,
|
| 1679 |
-
"grad_norm": 86024.0078125,
|
| 1680 |
-
"learning_rate": 4.067392784206944e-06,
|
| 1681 |
-
"loss": 1.8183,
|
| 1682 |
-
"step": 1195
|
| 1683 |
-
},
|
| 1684 |
{
|
| 1685 |
"epoch": 0.8168822328114363,
|
| 1686 |
-
"grad_norm":
|
| 1687 |
-
"learning_rate": 4.
|
| 1688 |
-
"loss": 1.
|
| 1689 |
"step": 1200
|
| 1690 |
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 0.820285908781484,
|
| 1693 |
-
"grad_norm": 112386.125,
|
| 1694 |
-
"learning_rate": 4.10142954390742e-06,
|
| 1695 |
-
"loss": 1.7165,
|
| 1696 |
-
"step": 1205
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 0.8236895847515316,
|
| 1700 |
-
"grad_norm": 108654.546875,
|
| 1701 |
-
"learning_rate": 4.118447923757659e-06,
|
| 1702 |
-
"loss": 1.8009,
|
| 1703 |
-
"step": 1210
|
| 1704 |
-
},
|
| 1705 |
-
{
|
| 1706 |
-
"epoch": 0.8270932607215793,
|
| 1707 |
-
"grad_norm": 98713.2109375,
|
| 1708 |
-
"learning_rate": 4.135466303607897e-06,
|
| 1709 |
-
"loss": 1.9296,
|
| 1710 |
-
"step": 1215
|
| 1711 |
-
},
|
| 1712 |
{
|
| 1713 |
"epoch": 0.8304969366916269,
|
| 1714 |
-
"grad_norm":
|
| 1715 |
-
"learning_rate": 4.
|
| 1716 |
-
"loss": 1.
|
| 1717 |
"step": 1220
|
| 1718 |
},
|
| 1719 |
-
{
|
| 1720 |
-
"epoch": 0.8339006126616746,
|
| 1721 |
-
"grad_norm": 198507.484375,
|
| 1722 |
-
"learning_rate": 4.169503063308374e-06,
|
| 1723 |
-
"loss": 1.7275,
|
| 1724 |
-
"step": 1225
|
| 1725 |
-
},
|
| 1726 |
-
{
|
| 1727 |
-
"epoch": 0.8373042886317222,
|
| 1728 |
-
"grad_norm": 248154.234375,
|
| 1729 |
-
"learning_rate": 4.186521443158612e-06,
|
| 1730 |
-
"loss": 1.9263,
|
| 1731 |
-
"step": 1230
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 0.8407079646017699,
|
| 1735 |
-
"grad_norm": 126586.7890625,
|
| 1736 |
-
"learning_rate": 4.20353982300885e-06,
|
| 1737 |
-
"loss": 1.8871,
|
| 1738 |
-
"step": 1235
|
| 1739 |
-
},
|
| 1740 |
{
|
| 1741 |
"epoch": 0.8441116405718175,
|
| 1742 |
-
"grad_norm":
|
| 1743 |
-
"learning_rate": 4.
|
| 1744 |
-
"loss": 1.
|
| 1745 |
"step": 1240
|
| 1746 |
},
|
| 1747 |
-
{
|
| 1748 |
-
"epoch": 0.8475153165418652,
|
| 1749 |
-
"grad_norm": 112431.09375,
|
| 1750 |
-
"learning_rate": 4.237576582709326e-06,
|
| 1751 |
-
"loss": 1.8617,
|
| 1752 |
-
"step": 1245
|
| 1753 |
-
},
|
| 1754 |
-
{
|
| 1755 |
-
"epoch": 0.8509189925119128,
|
| 1756 |
-
"grad_norm": 84201.109375,
|
| 1757 |
-
"learning_rate": 4.254594962559564e-06,
|
| 1758 |
-
"loss": 1.8836,
|
| 1759 |
-
"step": 1250
|
| 1760 |
-
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 0.8543226684819605,
|
| 1763 |
-
"grad_norm": 161661.96875,
|
| 1764 |
-
"learning_rate": 4.271613342409802e-06,
|
| 1765 |
-
"loss": 1.8717,
|
| 1766 |
-
"step": 1255
|
| 1767 |
-
},
|
| 1768 |
{
|
| 1769 |
"epoch": 0.8577263444520081,
|
| 1770 |
-
"grad_norm":
|
| 1771 |
-
"learning_rate": 4.
|
| 1772 |
-
"loss": 1.
|
| 1773 |
"step": 1260
|
| 1774 |
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 0.8611300204220558,
|
| 1777 |
-
"grad_norm": 131402.53125,
|
| 1778 |
-
"learning_rate": 4.305650102110279e-06,
|
| 1779 |
-
"loss": 1.7646,
|
| 1780 |
-
"step": 1265
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 0.8645336963921034,
|
| 1784 |
-
"grad_norm": 114878.9375,
|
| 1785 |
-
"learning_rate": 4.322668481960517e-06,
|
| 1786 |
-
"loss": 1.7294,
|
| 1787 |
-
"step": 1270
|
| 1788 |
-
},
|
| 1789 |
-
{
|
| 1790 |
-
"epoch": 0.8679373723621511,
|
| 1791 |
-
"grad_norm": 174311.171875,
|
| 1792 |
-
"learning_rate": 4.339686861810756e-06,
|
| 1793 |
-
"loss": 1.8443,
|
| 1794 |
-
"step": 1275
|
| 1795 |
-
},
|
| 1796 |
{
|
| 1797 |
"epoch": 0.8713410483321987,
|
| 1798 |
-
"grad_norm":
|
| 1799 |
-
"learning_rate": 4.
|
| 1800 |
-
"loss": 1.
|
| 1801 |
"step": 1280
|
| 1802 |
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 0.8747447243022465,
|
| 1805 |
-
"grad_norm": 134015.765625,
|
| 1806 |
-
"learning_rate": 4.373723621511233e-06,
|
| 1807 |
-
"loss": 1.7056,
|
| 1808 |
-
"step": 1285
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 0.8781484002722941,
|
| 1812 |
-
"grad_norm": 116424.71875,
|
| 1813 |
-
"learning_rate": 4.390742001361471e-06,
|
| 1814 |
-
"loss": 1.6381,
|
| 1815 |
-
"step": 1290
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 0.8815520762423418,
|
| 1819 |
-
"grad_norm": 114702.9140625,
|
| 1820 |
-
"learning_rate": 4.407760381211709e-06,
|
| 1821 |
-
"loss": 1.8646,
|
| 1822 |
-
"step": 1295
|
| 1823 |
-
},
|
| 1824 |
{
|
| 1825 |
"epoch": 0.8849557522123894,
|
| 1826 |
-
"grad_norm":
|
| 1827 |
-
"learning_rate": 4.
|
| 1828 |
-
"loss": 1.
|
| 1829 |
"step": 1300
|
| 1830 |
},
|
| 1831 |
-
{
|
| 1832 |
-
"epoch": 0.8883594281824371,
|
| 1833 |
-
"grad_norm": 70903.609375,
|
| 1834 |
-
"learning_rate": 4.441797140912186e-06,
|
| 1835 |
-
"loss": 1.8223,
|
| 1836 |
-
"step": 1305
|
| 1837 |
-
},
|
| 1838 |
-
{
|
| 1839 |
-
"epoch": 0.8917631041524847,
|
| 1840 |
-
"grad_norm": 163495.9375,
|
| 1841 |
-
"learning_rate": 4.458815520762424e-06,
|
| 1842 |
-
"loss": 1.6673,
|
| 1843 |
-
"step": 1310
|
| 1844 |
-
},
|
| 1845 |
-
{
|
| 1846 |
-
"epoch": 0.8951667801225324,
|
| 1847 |
-
"grad_norm": 122865.515625,
|
| 1848 |
-
"learning_rate": 4.4758339006126625e-06,
|
| 1849 |
-
"loss": 1.6363,
|
| 1850 |
-
"step": 1315
|
| 1851 |
-
},
|
| 1852 |
{
|
| 1853 |
"epoch": 0.89857045609258,
|
| 1854 |
-
"grad_norm":
|
| 1855 |
-
"learning_rate": 4.
|
| 1856 |
-
"loss": 1.
|
| 1857 |
"step": 1320
|
| 1858 |
},
|
| 1859 |
-
{
|
| 1860 |
-
"epoch": 0.9019741320626277,
|
| 1861 |
-
"grad_norm": 132914.34375,
|
| 1862 |
-
"learning_rate": 4.5098706603131384e-06,
|
| 1863 |
-
"loss": 1.6366,
|
| 1864 |
-
"step": 1325
|
| 1865 |
-
},
|
| 1866 |
-
{
|
| 1867 |
-
"epoch": 0.9053778080326753,
|
| 1868 |
-
"grad_norm": 243811.421875,
|
| 1869 |
-
"learning_rate": 4.526889040163377e-06,
|
| 1870 |
-
"loss": 1.7955,
|
| 1871 |
-
"step": 1330
|
| 1872 |
-
},
|
| 1873 |
-
{
|
| 1874 |
-
"epoch": 0.908781484002723,
|
| 1875 |
-
"grad_norm": 177361.71875,
|
| 1876 |
-
"learning_rate": 4.543907420013615e-06,
|
| 1877 |
-
"loss": 1.7663,
|
| 1878 |
-
"step": 1335
|
| 1879 |
-
},
|
| 1880 |
{
|
| 1881 |
"epoch": 0.9121851599727706,
|
| 1882 |
-
"grad_norm":
|
| 1883 |
-
"learning_rate": 4.
|
| 1884 |
-
"loss": 1.
|
| 1885 |
"step": 1340
|
| 1886 |
},
|
| 1887 |
-
{
|
| 1888 |
-
"epoch": 0.9155888359428183,
|
| 1889 |
-
"grad_norm": 72102.5546875,
|
| 1890 |
-
"learning_rate": 4.577944179714092e-06,
|
| 1891 |
-
"loss": 1.7733,
|
| 1892 |
-
"step": 1345
|
| 1893 |
-
},
|
| 1894 |
-
{
|
| 1895 |
-
"epoch": 0.9189925119128659,
|
| 1896 |
-
"grad_norm": 238094.09375,
|
| 1897 |
-
"learning_rate": 4.59496255956433e-06,
|
| 1898 |
-
"loss": 1.9633,
|
| 1899 |
-
"step": 1350
|
| 1900 |
-
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 0.9223961878829136,
|
| 1903 |
-
"grad_norm": 229956.09375,
|
| 1904 |
-
"learning_rate": 4.611980939414568e-06,
|
| 1905 |
-
"loss": 1.8553,
|
| 1906 |
-
"step": 1355
|
| 1907 |
-
},
|
| 1908 |
{
|
| 1909 |
"epoch": 0.9257998638529612,
|
| 1910 |
-
"grad_norm":
|
| 1911 |
-
"learning_rate": 4.
|
| 1912 |
-
"loss": 1.
|
| 1913 |
"step": 1360
|
| 1914 |
},
|
| 1915 |
-
{
|
| 1916 |
-
"epoch": 0.9292035398230089,
|
| 1917 |
-
"grad_norm": 84662.171875,
|
| 1918 |
-
"learning_rate": 4.646017699115045e-06,
|
| 1919 |
-
"loss": 1.757,
|
| 1920 |
-
"step": 1365
|
| 1921 |
-
},
|
| 1922 |
-
{
|
| 1923 |
-
"epoch": 0.9326072157930565,
|
| 1924 |
-
"grad_norm": 176475.6875,
|
| 1925 |
-
"learning_rate": 4.663036078965283e-06,
|
| 1926 |
-
"loss": 1.7104,
|
| 1927 |
-
"step": 1370
|
| 1928 |
-
},
|
| 1929 |
-
{
|
| 1930 |
-
"epoch": 0.9360108917631041,
|
| 1931 |
-
"grad_norm": 168853.984375,
|
| 1932 |
-
"learning_rate": 4.680054458815521e-06,
|
| 1933 |
-
"loss": 1.796,
|
| 1934 |
-
"step": 1375
|
| 1935 |
-
},
|
| 1936 |
{
|
| 1937 |
"epoch": 0.9394145677331518,
|
| 1938 |
-
"grad_norm":
|
| 1939 |
-
"learning_rate": 4.
|
| 1940 |
-
"loss": 1.
|
| 1941 |
"step": 1380
|
| 1942 |
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 0.9428182437031994,
|
| 1945 |
-
"grad_norm": 220900.59375,
|
| 1946 |
-
"learning_rate": 4.714091218515998e-06,
|
| 1947 |
-
"loss": 1.7126,
|
| 1948 |
-
"step": 1385
|
| 1949 |
-
},
|
| 1950 |
-
{
|
| 1951 |
-
"epoch": 0.9462219196732471,
|
| 1952 |
-
"grad_norm": 184930.40625,
|
| 1953 |
-
"learning_rate": 4.731109598366236e-06,
|
| 1954 |
-
"loss": 1.6739,
|
| 1955 |
-
"step": 1390
|
| 1956 |
-
},
|
| 1957 |
-
{
|
| 1958 |
-
"epoch": 0.9496255956432947,
|
| 1959 |
-
"grad_norm": 155985.140625,
|
| 1960 |
-
"learning_rate": 4.7481279782164745e-06,
|
| 1961 |
-
"loss": 1.7207,
|
| 1962 |
-
"step": 1395
|
| 1963 |
-
},
|
| 1964 |
{
|
| 1965 |
"epoch": 0.9530292716133424,
|
| 1966 |
-
"grad_norm":
|
| 1967 |
-
"learning_rate": 4.
|
| 1968 |
-
"loss": 1.
|
| 1969 |
"step": 1400
|
| 1970 |
},
|
| 1971 |
-
{
|
| 1972 |
-
"epoch": 0.95643294758339,
|
| 1973 |
-
"grad_norm": 201375.0625,
|
| 1974 |
-
"learning_rate": 4.7821647379169505e-06,
|
| 1975 |
-
"loss": 1.6343,
|
| 1976 |
-
"step": 1405
|
| 1977 |
-
},
|
| 1978 |
-
{
|
| 1979 |
-
"epoch": 0.9598366235534377,
|
| 1980 |
-
"grad_norm": 176404.828125,
|
| 1981 |
-
"learning_rate": 4.7991831177671884e-06,
|
| 1982 |
-
"loss": 1.8452,
|
| 1983 |
-
"step": 1410
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 0.9632402995234853,
|
| 1987 |
-
"grad_norm": 232695.171875,
|
| 1988 |
-
"learning_rate": 4.816201497617427e-06,
|
| 1989 |
-
"loss": 1.7085,
|
| 1990 |
-
"step": 1415
|
| 1991 |
-
},
|
| 1992 |
{
|
| 1993 |
"epoch": 0.966643975493533,
|
| 1994 |
-
"grad_norm":
|
| 1995 |
-
"learning_rate": 4.
|
| 1996 |
-
"loss": 1.
|
| 1997 |
"step": 1420
|
| 1998 |
},
|
| 1999 |
-
{
|
| 2000 |
-
"epoch": 0.9700476514635806,
|
| 2001 |
-
"grad_norm": 226258.3125,
|
| 2002 |
-
"learning_rate": 4.850238257317903e-06,
|
| 2003 |
-
"loss": 1.7416,
|
| 2004 |
-
"step": 1425
|
| 2005 |
-
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 0.9734513274336283,
|
| 2008 |
-
"grad_norm": 142705.59375,
|
| 2009 |
-
"learning_rate": 4.867256637168142e-06,
|
| 2010 |
-
"loss": 1.7345,
|
| 2011 |
-
"step": 1430
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 0.9768550034036759,
|
| 2015 |
-
"grad_norm": 187446.59375,
|
| 2016 |
-
"learning_rate": 4.88427501701838e-06,
|
| 2017 |
-
"loss": 1.7338,
|
| 2018 |
-
"step": 1435
|
| 2019 |
-
},
|
| 2020 |
{
|
| 2021 |
"epoch": 0.9802586793737236,
|
| 2022 |
-
"grad_norm":
|
| 2023 |
-
"learning_rate": 4.
|
| 2024 |
-
"loss": 1.
|
| 2025 |
"step": 1440
|
| 2026 |
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 0.9836623553437712,
|
| 2029 |
-
"grad_norm": 115865.0,
|
| 2030 |
-
"learning_rate": 4.918311776718857e-06,
|
| 2031 |
-
"loss": 1.9282,
|
| 2032 |
-
"step": 1445
|
| 2033 |
-
},
|
| 2034 |
-
{
|
| 2035 |
-
"epoch": 0.9870660313138189,
|
| 2036 |
-
"grad_norm": 130383.8125,
|
| 2037 |
-
"learning_rate": 4.935330156569095e-06,
|
| 2038 |
-
"loss": 1.6302,
|
| 2039 |
-
"step": 1450
|
| 2040 |
-
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 0.9904697072838666,
|
| 2043 |
-
"grad_norm": 145943.125,
|
| 2044 |
-
"learning_rate": 4.952348536419334e-06,
|
| 2045 |
-
"loss": 1.6837,
|
| 2046 |
-
"step": 1455
|
| 2047 |
-
},
|
| 2048 |
{
|
| 2049 |
"epoch": 0.9938733832539143,
|
| 2050 |
-
"grad_norm":
|
| 2051 |
-
"learning_rate": 4.
|
| 2052 |
-
"loss": 1.
|
| 2053 |
"step": 1460
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 0.9972770592239619,
|
| 2057 |
-
"grad_norm": 238323.84375,
|
| 2058 |
-
"learning_rate": 4.98638529611981e-06,
|
| 2059 |
-
"loss": 1.7442,
|
| 2060 |
-
"step": 1465
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 1.0,
|
| 2064 |
-
"eval_accuracy": 0.1981891348088531,
|
| 2065 |
-
"eval_loss": 1.6977837085723877,
|
| 2066 |
-
"eval_runtime": 314.1889,
|
| 2067 |
-
"eval_samples_per_second": 9.491,
|
| 2068 |
-
"eval_steps_per_second": 1.187,
|
| 2069 |
-
"step": 1469
|
| 2070 |
}
|
| 2071 |
],
|
| 2072 |
-
"logging_steps":
|
| 2073 |
"max_steps": 14690,
|
| 2074 |
"num_input_tokens_seen": 0,
|
| 2075 |
"num_train_epochs": 10,
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
"global_step": 1469,
|
|
|
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.013614703880190605,
|
| 13 |
+
"grad_norm": 46384.1953125,
|
| 14 |
+
"learning_rate": 1.360544217687075e-07,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
"loss": 2.0802,
|
| 16 |
+
"step": 20
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.02722940776038121,
|
| 20 |
+
"grad_norm": 35930.53515625,
|
| 21 |
+
"learning_rate": 2.72108843537415e-07,
|
| 22 |
+
"loss": 2.08,
|
| 23 |
"step": 40
|
| 24 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
{
|
| 26 |
"epoch": 0.04084411164057182,
|
| 27 |
+
"grad_norm": 23965.296875,
|
| 28 |
+
"learning_rate": 4.0816326530612243e-07,
|
| 29 |
+
"loss": 2.0796,
|
| 30 |
"step": 60
|
| 31 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
{
|
| 33 |
"epoch": 0.05445881552076242,
|
| 34 |
+
"grad_norm": 32476.51953125,
|
| 35 |
+
"learning_rate": 5.4421768707483e-07,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 36 |
"loss": 2.0791,
|
| 37 |
+
"step": 80
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.06807351940095303,
|
| 41 |
+
"grad_norm": 31751.640625,
|
| 42 |
+
"learning_rate": 6.802721088435376e-07,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
"loss": 2.0793,
|
| 44 |
+
"step": 100
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.08168822328114364,
|
| 48 |
+
"grad_norm": 29360.6328125,
|
| 49 |
+
"learning_rate": 8.163265306122449e-07,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
"loss": 2.0793,
|
| 51 |
+
"step": 120
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.09530292716133425,
|
| 55 |
+
"grad_norm": 31927.123046875,
|
| 56 |
+
"learning_rate": 9.523809523809525e-07,
|
| 57 |
+
"loss": 2.0788,
|
| 58 |
"step": 140
|
| 59 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
{
|
| 61 |
"epoch": 0.10891763104152484,
|
| 62 |
+
"grad_norm": 31685.216796875,
|
| 63 |
+
"learning_rate": 1.08843537414966e-06,
|
| 64 |
+
"loss": 2.0783,
|
| 65 |
"step": 160
|
| 66 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
{
|
| 68 |
"epoch": 0.12253233492171545,
|
| 69 |
+
"grad_norm": 24569.869140625,
|
| 70 |
+
"learning_rate": 1.2244897959183673e-06,
|
| 71 |
+
"loss": 2.078,
|
| 72 |
"step": 180
|
| 73 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
{
|
| 75 |
"epoch": 0.13614703880190607,
|
| 76 |
+
"grad_norm": 29527.107421875,
|
| 77 |
+
"learning_rate": 1.3605442176870751e-06,
|
| 78 |
+
"loss": 2.0772,
|
| 79 |
"step": 200
|
| 80 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
{
|
| 82 |
"epoch": 0.14976174268209666,
|
| 83 |
+
"grad_norm": 31908.814453125,
|
| 84 |
+
"learning_rate": 1.4965986394557825e-06,
|
| 85 |
+
"loss": 2.0768,
|
| 86 |
"step": 220
|
| 87 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
{
|
| 89 |
"epoch": 0.16337644656228728,
|
| 90 |
+
"grad_norm": 34667.72265625,
|
| 91 |
+
"learning_rate": 1.6326530612244897e-06,
|
| 92 |
+
"loss": 2.0754,
|
| 93 |
"step": 240
|
| 94 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 95 |
{
|
| 96 |
"epoch": 0.17699115044247787,
|
| 97 |
+
"grad_norm": 36386.16015625,
|
| 98 |
+
"learning_rate": 1.7687074829931975e-06,
|
| 99 |
+
"loss": 2.0742,
|
| 100 |
"step": 260
|
| 101 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 102 |
{
|
| 103 |
"epoch": 0.1906058543226685,
|
| 104 |
+
"grad_norm": 24842.310546875,
|
| 105 |
+
"learning_rate": 1.904761904761905e-06,
|
| 106 |
+
"loss": 2.0722,
|
| 107 |
"step": 280
|
| 108 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 109 |
{
|
| 110 |
"epoch": 0.2042205582028591,
|
| 111 |
+
"grad_norm": 28252.93359375,
|
| 112 |
+
"learning_rate": 2.0408163265306125e-06,
|
| 113 |
+
"loss": 2.0696,
|
| 114 |
"step": 300
|
| 115 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 116 |
{
|
| 117 |
"epoch": 0.21783526208304968,
|
| 118 |
+
"grad_norm": 18715.328125,
|
| 119 |
+
"learning_rate": 2.17687074829932e-06,
|
| 120 |
+
"loss": 2.0682,
|
| 121 |
"step": 320
|
| 122 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
{
|
| 124 |
"epoch": 0.2314499659632403,
|
| 125 |
+
"grad_norm": 34919.875,
|
| 126 |
+
"learning_rate": 2.3129251700680273e-06,
|
| 127 |
+
"loss": 2.0648,
|
| 128 |
"step": 340
|
| 129 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 130 |
{
|
| 131 |
"epoch": 0.2450646698434309,
|
| 132 |
+
"grad_norm": 41688.1484375,
|
| 133 |
+
"learning_rate": 2.4489795918367347e-06,
|
| 134 |
+
"loss": 2.0573,
|
| 135 |
"step": 360
|
| 136 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 137 |
{
|
| 138 |
"epoch": 0.2586793737236215,
|
| 139 |
+
"grad_norm": 38931.60546875,
|
| 140 |
+
"learning_rate": 2.5850340136054425e-06,
|
| 141 |
+
"loss": 2.052,
|
| 142 |
"step": 380
|
| 143 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 144 |
{
|
| 145 |
"epoch": 0.27229407760381213,
|
| 146 |
+
"grad_norm": 37335.18359375,
|
| 147 |
+
"learning_rate": 2.7210884353741503e-06,
|
| 148 |
+
"loss": 2.0462,
|
| 149 |
"step": 400
|
| 150 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 151 |
{
|
| 152 |
"epoch": 0.2859087814840027,
|
| 153 |
+
"grad_norm": 32101.99609375,
|
| 154 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 155 |
+
"loss": 2.0439,
|
| 156 |
"step": 420
|
| 157 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 158 |
{
|
| 159 |
"epoch": 0.2995234853641933,
|
| 160 |
+
"grad_norm": 44584.16796875,
|
| 161 |
+
"learning_rate": 2.993197278911565e-06,
|
| 162 |
+
"loss": 2.0354,
|
| 163 |
"step": 440
|
| 164 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 165 |
{
|
| 166 |
"epoch": 0.3131381892443839,
|
| 167 |
+
"grad_norm": 45670.23046875,
|
| 168 |
+
"learning_rate": 3.1292517006802725e-06,
|
| 169 |
+
"loss": 2.0189,
|
| 170 |
"step": 460
|
| 171 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 172 |
{
|
| 173 |
"epoch": 0.32675289312457456,
|
| 174 |
+
"grad_norm": 35136.80859375,
|
| 175 |
+
"learning_rate": 3.2653061224489794e-06,
|
| 176 |
+
"loss": 2.0123,
|
| 177 |
"step": 480
|
| 178 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 179 |
{
|
| 180 |
"epoch": 0.34036759700476515,
|
| 181 |
+
"grad_norm": 45886.94140625,
|
| 182 |
+
"learning_rate": 3.4013605442176872e-06,
|
| 183 |
+
"loss": 2.0115,
|
| 184 |
"step": 500
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.34036759700476515,
|
| 188 |
+
"eval_accuracy": 0.1981891348088531,
|
| 189 |
+
"eval_loss": 1.9869346618652344,
|
| 190 |
+
"eval_runtime": 306.1857,
|
| 191 |
+
"eval_samples_per_second": 9.739,
|
| 192 |
+
"eval_steps_per_second": 1.218,
|
| 193 |
+
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 194 |
},
|
| 195 |
{
|
| 196 |
"epoch": 0.35398230088495575,
|
| 197 |
+
"grad_norm": 51081.21875,
|
| 198 |
+
"learning_rate": 3.537414965986395e-06,
|
| 199 |
+
"loss": 1.9899,
|
| 200 |
"step": 520
|
| 201 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 202 |
{
|
| 203 |
"epoch": 0.36759700476514634,
|
| 204 |
+
"grad_norm": 65917.5625,
|
| 205 |
+
"learning_rate": 3.6734693877551024e-06,
|
| 206 |
+
"loss": 1.9912,
|
| 207 |
"step": 540
|
| 208 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 209 |
{
|
| 210 |
"epoch": 0.381211708645337,
|
| 211 |
+
"grad_norm": 70819.2109375,
|
| 212 |
+
"learning_rate": 3.80952380952381e-06,
|
| 213 |
+
"loss": 1.9697,
|
| 214 |
"step": 560
|
| 215 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 216 |
{
|
| 217 |
"epoch": 0.3948264125255276,
|
| 218 |
+
"grad_norm": 34979.16015625,
|
| 219 |
+
"learning_rate": 3.945578231292517e-06,
|
| 220 |
+
"loss": 1.9669,
|
| 221 |
"step": 580
|
| 222 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 223 |
{
|
| 224 |
"epoch": 0.4084411164057182,
|
| 225 |
+
"grad_norm": 86509.28125,
|
| 226 |
+
"learning_rate": 4.081632653061225e-06,
|
| 227 |
+
"loss": 1.9561,
|
| 228 |
"step": 600
|
| 229 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 230 |
{
|
| 231 |
"epoch": 0.42205582028590877,
|
| 232 |
+
"grad_norm": 80961.046875,
|
| 233 |
+
"learning_rate": 4.217687074829933e-06,
|
| 234 |
+
"loss": 1.9278,
|
| 235 |
"step": 620
|
| 236 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 237 |
{
|
| 238 |
"epoch": 0.43567052416609936,
|
| 239 |
+
"grad_norm": 136664.796875,
|
| 240 |
+
"learning_rate": 4.35374149659864e-06,
|
| 241 |
+
"loss": 1.9133,
|
| 242 |
"step": 640
|
| 243 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 244 |
{
|
| 245 |
"epoch": 0.44928522804629,
|
| 246 |
+
"grad_norm": 82047.3828125,
|
| 247 |
+
"learning_rate": 4.489795918367348e-06,
|
| 248 |
+
"loss": 1.8824,
|
| 249 |
"step": 660
|
| 250 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 251 |
{
|
| 252 |
"epoch": 0.4628999319264806,
|
| 253 |
+
"grad_norm": 100084.4375,
|
| 254 |
+
"learning_rate": 4.6258503401360546e-06,
|
| 255 |
+
"loss": 1.8542,
|
| 256 |
"step": 680
|
| 257 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 258 |
{
|
| 259 |
"epoch": 0.4765146358066712,
|
| 260 |
+
"grad_norm": 91592.125,
|
| 261 |
+
"learning_rate": 4.761904761904762e-06,
|
| 262 |
+
"loss": 1.8855,
|
| 263 |
"step": 700
|
| 264 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 265 |
{
|
| 266 |
"epoch": 0.4901293396868618,
|
| 267 |
+
"grad_norm": 107294.0546875,
|
| 268 |
+
"learning_rate": 4.897959183673469e-06,
|
| 269 |
+
"loss": 1.869,
|
| 270 |
"step": 720
|
| 271 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 272 |
{
|
| 273 |
"epoch": 0.5037440435670524,
|
| 274 |
+
"grad_norm": 65171.02734375,
|
| 275 |
+
"learning_rate": 4.998208527409531e-06,
|
| 276 |
+
"loss": 1.8418,
|
| 277 |
"step": 740
|
| 278 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 279 |
{
|
| 280 |
"epoch": 0.517358747447243,
|
| 281 |
+
"grad_norm": 122938.7578125,
|
| 282 |
+
"learning_rate": 4.991042637047653e-06,
|
| 283 |
+
"loss": 1.7756,
|
| 284 |
"step": 760
|
| 285 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 286 |
{
|
| 287 |
"epoch": 0.5309734513274337,
|
| 288 |
+
"grad_norm": 149156.90625,
|
| 289 |
+
"learning_rate": 4.9838767466857765e-06,
|
| 290 |
+
"loss": 1.7877,
|
| 291 |
"step": 780
|
| 292 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 293 |
{
|
| 294 |
"epoch": 0.5445881552076243,
|
| 295 |
+
"grad_norm": 114232.359375,
|
| 296 |
+
"learning_rate": 4.976710856323898e-06,
|
| 297 |
+
"loss": 1.7956,
|
| 298 |
"step": 800
|
| 299 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 300 |
{
|
| 301 |
"epoch": 0.5582028590878149,
|
| 302 |
+
"grad_norm": 71295.1171875,
|
| 303 |
+
"learning_rate": 4.969544965962021e-06,
|
| 304 |
+
"loss": 1.8045,
|
| 305 |
"step": 820
|
| 306 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 307 |
{
|
| 308 |
"epoch": 0.5718175629680055,
|
| 309 |
+
"grad_norm": 107194.9453125,
|
| 310 |
+
"learning_rate": 4.962379075600144e-06,
|
| 311 |
+
"loss": 1.7672,
|
| 312 |
"step": 840
|
| 313 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 314 |
{
|
| 315 |
"epoch": 0.585432266848196,
|
| 316 |
+
"grad_norm": 187415.6875,
|
| 317 |
+
"learning_rate": 4.9552131852382664e-06,
|
| 318 |
+
"loss": 1.7479,
|
| 319 |
"step": 860
|
| 320 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 321 |
{
|
| 322 |
"epoch": 0.5990469707283866,
|
| 323 |
+
"grad_norm": 138833.421875,
|
| 324 |
+
"learning_rate": 4.948047294876388e-06,
|
| 325 |
+
"loss": 1.8075,
|
| 326 |
"step": 880
|
| 327 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 328 |
{
|
| 329 |
"epoch": 0.6126616746085772,
|
| 330 |
+
"grad_norm": 164837.234375,
|
| 331 |
+
"learning_rate": 4.940881404514512e-06,
|
| 332 |
+
"loss": 1.7473,
|
| 333 |
"step": 900
|
| 334 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 335 |
{
|
| 336 |
"epoch": 0.6262763784887678,
|
| 337 |
+
"grad_norm": 148592.34375,
|
| 338 |
+
"learning_rate": 4.933715514152634e-06,
|
| 339 |
+
"loss": 1.647,
|
| 340 |
"step": 920
|
| 341 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 342 |
{
|
| 343 |
"epoch": 0.6398910823689585,
|
| 344 |
+
"grad_norm": 219803.328125,
|
| 345 |
+
"learning_rate": 4.926549623790756e-06,
|
| 346 |
+
"loss": 1.7584,
|
| 347 |
"step": 940
|
| 348 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 349 |
{
|
| 350 |
"epoch": 0.6535057862491491,
|
| 351 |
+
"grad_norm": 65507.84375,
|
| 352 |
+
"learning_rate": 4.919383733428879e-06,
|
| 353 |
+
"loss": 1.7161,
|
| 354 |
"step": 960
|
| 355 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 356 |
{
|
| 357 |
"epoch": 0.6671204901293397,
|
| 358 |
+
"grad_norm": 216308.921875,
|
| 359 |
+
"learning_rate": 4.912217843067002e-06,
|
| 360 |
+
"loss": 1.787,
|
| 361 |
"step": 980
|
| 362 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 363 |
{
|
| 364 |
"epoch": 0.6807351940095303,
|
| 365 |
+
"grad_norm": 169682.25,
|
| 366 |
+
"learning_rate": 4.905051952705124e-06,
|
| 367 |
+
"loss": 1.7747,
|
| 368 |
"step": 1000
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"epoch": 0.6807351940095303,
|
| 372 |
+
"eval_accuracy": 0.1981891348088531,
|
| 373 |
+
"eval_loss": 1.6881698369979858,
|
| 374 |
+
"eval_runtime": 294.5683,
|
| 375 |
+
"eval_samples_per_second": 10.123,
|
| 376 |
+
"eval_steps_per_second": 1.266,
|
| 377 |
+
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 378 |
},
|
| 379 |
{
|
| 380 |
"epoch": 0.6943498978897209,
|
| 381 |
+
"grad_norm": 247486.65625,
|
| 382 |
+
"learning_rate": 4.897886062343246e-06,
|
| 383 |
+
"loss": 1.765,
|
| 384 |
"step": 1020
|
| 385 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 386 |
{
|
| 387 |
"epoch": 0.7079646017699115,
|
| 388 |
+
"grad_norm": 169864.59375,
|
| 389 |
+
"learning_rate": 4.890720171981369e-06,
|
| 390 |
+
"loss": 1.7452,
|
| 391 |
"step": 1040
|
| 392 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 393 |
{
|
| 394 |
"epoch": 0.7215793056501021,
|
| 395 |
+
"grad_norm": 141110.296875,
|
| 396 |
+
"learning_rate": 4.883554281619492e-06,
|
| 397 |
+
"loss": 1.8073,
|
| 398 |
"step": 1060
|
| 399 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 400 |
{
|
| 401 |
"epoch": 0.7351940095302927,
|
| 402 |
+
"grad_norm": 168530.28125,
|
| 403 |
+
"learning_rate": 4.876388391257614e-06,
|
| 404 |
+
"loss": 1.7331,
|
| 405 |
"step": 1080
|
| 406 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 407 |
{
|
| 408 |
"epoch": 0.7488087134104833,
|
| 409 |
+
"grad_norm": 131218.578125,
|
| 410 |
+
"learning_rate": 4.869222500895736e-06,
|
| 411 |
+
"loss": 1.6542,
|
| 412 |
"step": 1100
|
| 413 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 414 |
{
|
| 415 |
"epoch": 0.762423417290674,
|
| 416 |
+
"grad_norm": 151837.5,
|
| 417 |
+
"learning_rate": 4.862056610533859e-06,
|
| 418 |
+
"loss": 1.7534,
|
| 419 |
"step": 1120
|
| 420 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 421 |
{
|
| 422 |
"epoch": 0.7760381211708646,
|
| 423 |
+
"grad_norm": 131823.84375,
|
| 424 |
+
"learning_rate": 4.854890720171982e-06,
|
| 425 |
+
"loss": 1.7498,
|
| 426 |
"step": 1140
|
| 427 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 428 |
{
|
| 429 |
"epoch": 0.7896528250510552,
|
| 430 |
+
"grad_norm": 219690.40625,
|
| 431 |
+
"learning_rate": 4.847724829810104e-06,
|
| 432 |
+
"loss": 1.643,
|
| 433 |
"step": 1160
|
| 434 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 435 |
{
|
| 436 |
"epoch": 0.8032675289312458,
|
| 437 |
+
"grad_norm": 121273.3125,
|
| 438 |
+
"learning_rate": 4.840558939448226e-06,
|
| 439 |
+
"loss": 1.7032,
|
| 440 |
"step": 1180
|
| 441 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 442 |
{
|
| 443 |
"epoch": 0.8168822328114363,
|
| 444 |
+
"grad_norm": 204578.28125,
|
| 445 |
+
"learning_rate": 4.83339304908635e-06,
|
| 446 |
+
"loss": 1.7133,
|
| 447 |
"step": 1200
|
| 448 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 449 |
{
|
| 450 |
"epoch": 0.8304969366916269,
|
| 451 |
+
"grad_norm": 87904.203125,
|
| 452 |
+
"learning_rate": 4.826227158724472e-06,
|
| 453 |
+
"loss": 1.7369,
|
| 454 |
"step": 1220
|
| 455 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 456 |
{
|
| 457 |
"epoch": 0.8441116405718175,
|
| 458 |
+
"grad_norm": 94867.0625,
|
| 459 |
+
"learning_rate": 4.819061268362594e-06,
|
| 460 |
+
"loss": 1.8071,
|
| 461 |
"step": 1240
|
| 462 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 463 |
{
|
| 464 |
"epoch": 0.8577263444520081,
|
| 465 |
+
"grad_norm": 231554.109375,
|
| 466 |
+
"learning_rate": 4.811895378000717e-06,
|
| 467 |
+
"loss": 1.8353,
|
| 468 |
"step": 1260
|
| 469 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 470 |
{
|
| 471 |
"epoch": 0.8713410483321987,
|
| 472 |
+
"grad_norm": 178638.8125,
|
| 473 |
+
"learning_rate": 4.80472948763884e-06,
|
| 474 |
+
"loss": 1.7133,
|
| 475 |
"step": 1280
|
| 476 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 477 |
{
|
| 478 |
"epoch": 0.8849557522123894,
|
| 479 |
+
"grad_norm": 166282.265625,
|
| 480 |
+
"learning_rate": 4.7975635972769616e-06,
|
| 481 |
+
"loss": 1.6657,
|
| 482 |
"step": 1300
|
| 483 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 484 |
{
|
| 485 |
"epoch": 0.89857045609258,
|
| 486 |
+
"grad_norm": 192489.40625,
|
| 487 |
+
"learning_rate": 4.790397706915085e-06,
|
| 488 |
+
"loss": 1.6692,
|
| 489 |
"step": 1320
|
| 490 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 491 |
{
|
| 492 |
"epoch": 0.9121851599727706,
|
| 493 |
+
"grad_norm": 252441.734375,
|
| 494 |
+
"learning_rate": 4.783231816553207e-06,
|
| 495 |
+
"loss": 1.6666,
|
| 496 |
"step": 1340
|
| 497 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 498 |
{
|
| 499 |
"epoch": 0.9257998638529612,
|
| 500 |
+
"grad_norm": 134902.515625,
|
| 501 |
+
"learning_rate": 4.77606592619133e-06,
|
| 502 |
+
"loss": 1.8356,
|
| 503 |
"step": 1360
|
| 504 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 505 |
{
|
| 506 |
"epoch": 0.9394145677331518,
|
| 507 |
+
"grad_norm": 207219.609375,
|
| 508 |
+
"learning_rate": 4.768900035829452e-06,
|
| 509 |
+
"loss": 1.7977,
|
| 510 |
"step": 1380
|
| 511 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 512 |
{
|
| 513 |
"epoch": 0.9530292716133424,
|
| 514 |
+
"grad_norm": 110303.21875,
|
| 515 |
+
"learning_rate": 4.761734145467575e-06,
|
| 516 |
+
"loss": 1.6795,
|
| 517 |
"step": 1400
|
| 518 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 519 |
{
|
| 520 |
"epoch": 0.966643975493533,
|
| 521 |
+
"grad_norm": 155455.640625,
|
| 522 |
+
"learning_rate": 4.754568255105697e-06,
|
| 523 |
+
"loss": 1.7061,
|
| 524 |
"step": 1420
|
| 525 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 526 |
{
|
| 527 |
"epoch": 0.9802586793737236,
|
| 528 |
+
"grad_norm": 238588.40625,
|
| 529 |
+
"learning_rate": 4.74740236474382e-06,
|
| 530 |
+
"loss": 1.7127,
|
| 531 |
"step": 1440
|
| 532 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 533 |
{
|
| 534 |
"epoch": 0.9938733832539143,
|
| 535 |
+
"grad_norm": 358719.46875,
|
| 536 |
+
"learning_rate": 4.740236474381942e-06,
|
| 537 |
+
"loss": 1.7185,
|
| 538 |
"step": 1460
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 539 |
}
|
| 540 |
],
|
| 541 |
+
"logging_steps": 20,
|
| 542 |
"max_steps": 14690,
|
| 543 |
"num_input_tokens_seen": 0,
|
| 544 |
"num_train_epochs": 10,
|
checkpoint-1469/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0eecac33aad667da4f4ae5a58f5c275e5b47b428ee45b4e90760cd1ea6b1c435
|
| 3 |
size 5176
|